What's new in OpenClaw — new features, fixes, and breaking changes explained
v2026.7.2-beta.1
📅 July 15, 2026
⬇️148 downloads❤️9 reactions
🆕 New Features
**Remote coding sessions:** run Control UI sessions on cloud workers, open Codex and Claude catalog sessions in terminals on their owning hosts, and resume OpenCode and Pi sessions directly in a terminal.
**Native automation and nodes:** bring Automations parity to mobile, add foreground Voice Wake on Android, and expose camera, location, and notification capabilities from headless Linux nodes.
**Safer channel operation:** **Fixed:** Telegram durable-ingress loss after restarts, keep Signal stop and approval controls responsive during active turns, and stop channel allowlists from granting owner access. Thanks @obviyus, @arduano, and @yetval.
**Guided Control UI setup:** configure model providers from Settings, onboard channels through a guided setup page, and choose images and models while creating sessions. Thanks @alexandre-leng and @fuller-stack-dev.
**Gateway and session recovery:** **Fixed:** Restart admission from wedging the Gateway, recover reply sessions after finalization stalls, and keep one-shot cron jobs enabled through lifecycle claim races. Thanks @obviyus, @joshavant, @charliemeyer2000, and @SL4N.
**Install and packaging:** add Linux deb and AppImage bundles with Gateway guidance, publish them from stable main-based releases, and let Windows installs continue immediately after winget adds Node.js.
**Cloud workers:** add session placement, dispatch, and worker-turn routing for remote session execution.
**Paired-node coding agents:** discover OpenCode and Pi sessions and continue Codex and Claude catalog sessions through streaming CLI agent runs.
**Catalog sessions:** create eligible catalog sessions directly from the Control UI sidebar.
**Managed worktrees:** configure cleanup limits by count and total size from Settings.
...and 10 more
🐛 Bug Fixes
**Terminal sessions:** preserve early keystrokes, Enabled paired-node terminal actions, authorize paired-node catalog reads, recover after stale Control UI connections, and keep slow clients attached under heavy output. Thanks @vincentkoc and @zw-xysk.
**Control UI configuration:** auto-save validated edits, restore autosave gates, show unapplied-restart state, and keep normal content scrolling and selection behavior.
**Signal recovery:** reconnect after stalled container handshakes, **Fixed:** Overlapping daemons during restart, and preserve username targets during phone normalization. Thanks @hugenshen, @ZHOUKAILIAN, @UberKitten, and @ly-wang19.
**Authorization boundaries:** require operator-admin authority for paired-node directory browsing and block prototype pollution in migration config merges. Thanks @yangxiansheng, @joshavant, @yetval, and @ruel225.
**Conversation identity:** retain speaker attribution across group turns and reduce malformed Discord guild reply tool calls.
**Provider setup:** show OpenRouter OAuth denial errors, find credentials stored under auth aliases, and keep unresolved API-key providers visible for repair. Thanks @VectorPeak and @altaywtf.
**Claude and Codex sessions:** keep Claude turns open for native background work, **Fixed:** Codex shutdown warnings after live updates, complete Codex onboarding login handoff, and retire completed hook relays. Thanks @obviyus, @fuller-stack-dev, and @bek91.
**Apple apps:** connect iPhone and Watch clients to protocol-v3 Gateways, reject incomplete attachment shares, and honor the pinned Swift toolchain in iOS release archives. Thanks @jincheng-xydt, @harjothkhara, and @joshavant.
**Memory reliability:** preserve UTF-8 across split QMD output, keep canonical cache rows during legacy migration, and accept leading-zero JSON content lengths. Thanks @wahaha1223, @felirami, @Tony-ooo, @qingminlong, @vincentkoc, and @zw-xysk.
**Skills and Workshop:** keep globally installed ClawHub skills tracked, clean orphaned usage rows, surface verification errors, and default Workshop approvals to automatic handling. Thanks @ml12580, @mnowrot, @SunnyShu0925, and @shakkernerd.
**Control UI overhaul:** conversations are easier to organize and work with side by side, with live Tasks, clearer chat controls, **Improved:** Usage and cost views, files, downloads, pairing, approvals, and Gateway health kept close to the conversation.
**Easier setup from install to first chat:** guided setup provides clearer steps toward a working first chat, checks connections before saving them, and preserves earlier choices when setup is interrupted.
**Official app updates:** iOS and iPadOS, Android, and macOS received substantial work across setup, navigation, chat, voice, permissions, localization, files, scheduled work, offline reading, queued sends, connection recovery, and native session controls.
**Models and providers:** GPT-5.6 compatibility improves across supported OpenAI and Codex routes, Tencent Hy3 gains a complete setup path, Meta Model API adds Muse Spark 1.1, and broader Claude, Ollama, ClawRouter, LongCat, and provider work expands choice and reliability.
**Codex and connected coding agents:** `openclaw attach` gives Claude Code temporary access to a selected session, Codex delegation and native subagents return tracked results more reliably, Copilot gains broader provider choices, and long-running sessions and goals are easier to resume.
**Telegram:** live progress, photos and documents, topics, commands, retries, account routing, setup, and delivery all received substantial work.
**Slack:** threads, cards, progress, identity, reactions, and duplicate prevention improve, while longer conversations **Fixed:** More unnecessary waits.
**Discord:** replies, attachments, voice sessions, progress, reconnects, and multi-account behavior improve, with **Improved:** Unread cues and recovery from repeated session-resume failures.
**Apple Messages:** replies, typing, media, routing, setup guidance, and chat continuity **Improved:** Across supported bridge setups.
**Gateway crashes stop restarting forever:** repeatedly failing Gateways now leave a stable **Fixed:** Path instead of restarting indefinitely.
...and 8 more
🐛 Bug Fixes
**Chat delivery and replies:** messages, progress updates, quoted replies, media, and final results stay tied to the intended conversation more reliably across reconnects, restarts, retries, and channel-specific limits.
**Control UI and native apps:** fixes cover session state, stale completions, responsive layouts, scrolling, focus, accessibility, localization, pairing, voice, offline chat, and mobile recovery.
**Gateway and provider reliability:** startup, restart, migration, authentication, fallback, streaming, model discovery, and oversized-response failures now **Fixed:** Or recover more cleanly.
**Sessions and memory:** transcript, compaction, checkpoint, cache, rotation, search, and persistence fixes reduce lost state, stale views, and cross-session confusion.
**Browser, tools, and automation:** browser actions, downloads, fetches, searches, scheduled jobs, terminals, subprocesses, and MCP calls now handle cancellation, malformed input, timeouts, and broken streams more safely.
**Setup, updates, and packaging:** installers, containers, plugin updates, runtime checks, temporary files, and platform-specific launch paths receive recovery and compatibility fixes.
**Security and credentials:** additional hardening protects tokens, secrets, approval scopes, local paths, network destinations, attachments, archives, and untrusted provider or plugin responses.
**New models and providers:** add Featherless, Claude Sonnet 5 and Mythos 5, Meta Muse Spark 1.1, and ClawRouter; make GPT-5.6 the new-setup default with `/think ultra` for Sol and Terra plus `max` for Luna, honor Z.AI `max`, and refresh model availability after OAuth renewal. Thanks @vincentkoc, @vortexopenclaw, @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, @steipete-oai, @bdjben, @obviyus, @sallyom, @anyech, @100yenadmin, and @fuller-stack-dev.
**Control UI and native macOS chat:** make sessions primary with a minimal searchable sidebar, compact context ring, reasoning-effort slider, cleaner dashboard chrome, and a native macOS session browser with model and thinking pickers, slash commands, transcript export, and context usage.
**Conversational onboarding:** Crestodian now runs a real agent-loop setup across the CLI, web install, and macOS app, with AI-guided provider setup, model-judged approvals bound to exact operations, masked credential prompts, and deterministic fallback when no model is available.
**Offline and spoken mobile chat:** iOS and Android now pre-paint bounded per-gateway session and transcript caches offline; Apple Watch gains full voice turns, and iOS can speak replies through configured Gateway TTS with on-device fallback.
**Session organization and generated titles:** generate concise session titles through utility-model routing and add Gateway-backed groups, unread state, rename, fork, archive, and delete controls, plus group management across web, iOS, and Android. Thanks @zhangguiping-xydt and @Juliangsm.
**Telegram and Codex continuity:** pair Codex through private `/login`, steer active runs with `/steer` and `/tell`, preserve messages when transcript acknowledgement is missing, adopt durable turns safely, and recover final sends across Telegram formatting and flood-wait failures. Thanks @100yenadmin, @Kyzcreig, @obviyus, and @jalehman.
**Startup and upgrade recovery:** run container migrations before Gateway readiness, **Fixed:** Recoverable legacy state from blocking startup, and enter control-plane-safe mode after repeated unclean boots instead of restart flapping. Fixes #98565 Thanks @sallyom, @jacobtomlinson, @bdjben, @obviyus, and @shakkernerd.
**Meta provider:** **Bundled `muse-spark-1.1` Responses API** is now supported as a provider with streaming, tool calls, encrypted reasoning replay, onboarding, exact-model live validation, and standalone npm/ClawHub distribution as `@openclaw/meta-provider`. Thanks @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, and @vincentkoc.
**Android chat agent selector:** switch the active agent directly from the live chat screen while keeping chat, Talk mode, and home canvas on the same canonical session. Thanks @bcperry and @joshavant.
**Gateway host status:** show the connected Gateway's host, network address, OS, runtime, uptime, CPU, memory, and disk details in Control UI Settings.
...and 27 more
🐛 Bug Fixes
**Codex runtime switching:** accept the bundled Codex runtime for both `codex/*` and `openai/*` model routes while keeping unsupported provider/runtime pairs rejected.
**LM Studio and xAI repaired tool calls:** emit the terminal tool-call event before completion when promoting serialized plain-text tool calls, preserving the complete streaming lifecycle for consumers.
**Browser actions on Node 24:** keep browser request cancellation bound to the client and response lifetime instead of Node 24.16+'s prematurely aborted body-stream signal, preventing valid POST actions from failing after JSON parsing. Thanks @obviyus and @vincentkoc.
**SecretRef model credentials:** keep resolved provider secrets behind process-local sentinels through auth storage, stream setup, SDK configuration, and managed local-provider probing, then inject plaintext only at the final network or provider-plugin boundary while retaining exact-value log redaction.
**Telegram token redaction:** redact bot tokens even when log transports split them across chunks, preventing fragmented credentials from escaping structured and streamed logs.
**Telegram durable turn adoption:** complete spooled updates only after durable agent adoption, detach adopted turns from the ingress reply fence, and suppress superseded replies so recovery cannot leak stale or duplicate responses. Thanks @obviyus and @vincentkoc.
**Task delivery recovery:** normalize legacy delivery statuses while restoring task records so older queued work becomes runnable instead of aborting registry recovery. Fixes #103168 Thanks @bek91, @theo674, and @obviyus.
**Diagnostics provider evidence:** emit one deduplicated `provider.request` timeline event for every completed or failed model call, so opt-in timelines can prove real provider traffic without double-counting terminal paths. Fixes #103063.
**Lean local model shell access:** keep `exec` directly visible beside the default structured Tool Search controls so coding-tuned local models can use their shell fallback instead of searching for missing domain tools. Thanks @vincentkoc and @maweibin.
**OAuth refresh contention diagnostics:** keep local lock paths out of user-facing refresh failures and **Fixed:** Duplicate failure prefixes while preserving structured provider and profile classification.
**New models and providers:** add Featherless, Claude Sonnet 5 and Mythos 5, Meta Muse Spark 1.1, and ClawRouter; make GPT-5.6 the new-setup default with `/think ultra` for Sol and Terra plus `max` for Luna, honor Z.AI `max`, and refresh model availability after OAuth renewal. Thanks @vincentkoc, @vortexopenclaw, @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, @steipete-oai, @bdjben, @obviyus, @sallyom, @anyech, @100yenadmin, and @fuller-stack-dev.
**Control UI and native macOS chat:** make sessions primary with a minimal searchable sidebar, compact context ring, reasoning-effort slider, cleaner dashboard chrome, and a native macOS session browser with model and thinking pickers, slash commands, transcript export, and context usage.
**Conversational onboarding:** Crestodian now runs a real agent-loop setup across the CLI, web install, and macOS app, with AI-guided provider setup, model-judged approvals bound to exact operations, masked credential prompts, and deterministic fallback when no model is available.
**Offline and spoken mobile chat:** iOS and Android now pre-paint bounded per-gateway session and transcript caches offline; Apple Watch gains full voice turns, and iOS can speak replies through configured Gateway TTS with on-device fallback.
**Session organization and generated titles:** generate concise session titles through utility-model routing and add Gateway-backed groups, unread state, rename, fork, archive, and delete controls, plus group management across web, iOS, and Android. Thanks @zhangguiping-xydt and @Juliangsm.
**Telegram and Codex continuity:** pair Codex through private `/login`, steer active runs with `/steer` and `/tell`, preserve messages when transcript acknowledgement is missing, adopt durable turns safely, and recover final sends across Telegram formatting and flood-wait failures. Thanks @100yenadmin, @Kyzcreig, @obviyus, and @jalehman.
**Startup and upgrade recovery:** run container migrations before Gateway readiness, **Fixed:** Recoverable legacy state from blocking startup, and enter control-plane-safe mode after repeated unclean boots instead of restart flapping. Fixes #98565 Thanks @sallyom, @jacobtomlinson, @bdjben, @obviyus, and @shakkernerd.
**Meta provider:** **Bundled `muse-spark-1.1` Responses API** is now supported as a provider with streaming, tool calls, encrypted reasoning replay, onboarding, exact-model live validation, and standalone npm/ClawHub distribution as `@openclaw/meta-provider`. Thanks @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, and @vincentkoc.
**Android chat agent selector:** switch the active agent directly from the live chat screen while keeping chat, Talk mode, and home canvas on the same canonical session. Thanks @bcperry and @joshavant.
**Gateway host status:** show the connected Gateway's host, network address, OS, runtime, uptime, CPU, memory, and disk details in Control UI Settings.
...and 27 more
🐛 Bug Fixes
**Codex runtime switching:** accept the bundled Codex runtime for both `codex/*` and `openai/*` model routes while keeping unsupported provider/runtime pairs rejected.
**LM Studio and xAI repaired tool calls:** emit the terminal tool-call event before completion when promoting serialized plain-text tool calls, preserving the complete streaming lifecycle for consumers.
**Browser actions on Node 24:** keep browser request cancellation bound to the client and response lifetime instead of Node 24.16+'s prematurely aborted body-stream signal, preventing valid POST actions from failing after JSON parsing. Thanks @obviyus and @vincentkoc.
**SecretRef model credentials:** keep resolved provider secrets behind process-local sentinels through auth storage, stream setup, SDK configuration, and managed local-provider probing, then inject plaintext only at the final network or provider-plugin boundary while retaining exact-value log redaction.
**Telegram token redaction:** redact bot tokens even when log transports split them across chunks, preventing fragmented credentials from escaping structured and streamed logs.
**Telegram durable turn adoption:** complete spooled updates only after durable agent adoption, detach adopted turns from the ingress reply fence, and suppress superseded replies so recovery cannot leak stale or duplicate responses. Thanks @obviyus and @vincentkoc.
**Task delivery recovery:** normalize legacy delivery statuses while restoring task records so older queued work becomes runnable instead of aborting registry recovery. Fixes #103168 Thanks @bek91, @theo674, and @obviyus.
**Diagnostics provider evidence:** emit one deduplicated `provider.request` timeline event for every completed or failed model call, so opt-in timelines can prove real provider traffic without double-counting terminal paths. Fixes #103063.
**Lean local model shell access:** keep `exec` directly visible beside the default structured Tool Search controls so coding-tuned local models can use their shell fallback instead of searching for missing domain tools. Thanks @vincentkoc and @maweibin.
**OAuth refresh contention diagnostics:** keep local lock paths out of user-facing refresh failures and **Fixed:** Duplicate failure prefixes while preserving structured provider and profile classification.
**New models and providers:** add Featherless, Claude Sonnet 5 and Mythos 5, Meta Muse Spark 1.1, and ClawRouter; make GPT-5.6 the new-setup default with `/think ultra` for Sol and Terra plus `max` for Luna, honor Z.AI `max`, and refresh model availability after OAuth renewal. Thanks @vincentkoc, @vortexopenclaw, @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, @steipete-oai, @bdjben, @obviyus, @sallyom, @anyech, @100yenadmin, and @fuller-stack-dev.
**Control UI and native macOS chat:** make sessions primary with a minimal searchable sidebar, compact context ring, reasoning-effort slider, cleaner dashboard chrome, and a native macOS session browser with model and thinking pickers, slash commands, transcript export, and context usage.
**Conversational onboarding:** Crestodian now runs a real agent-loop setup across the CLI, web install, and macOS app, with AI-guided provider setup, model-judged approvals bound to exact operations, masked credential prompts, and deterministic fallback when no model is available.
**Offline and spoken mobile chat:** iOS and Android now pre-paint bounded per-gateway session and transcript caches offline; Apple Watch gains full voice turns, and iOS can speak replies through configured Gateway TTS with on-device fallback.
**Session organization and generated titles:** generate concise session titles through utility-model routing and add Gateway-backed groups, unread state, rename, fork, archive, and delete controls, plus group management across web, iOS, and Android. Thanks @zhangguiping-xydt and @Juliangsm.
**Telegram and Codex continuity:** pair Codex through private `/login`, steer active runs with `/steer` and `/tell`, preserve messages when transcript acknowledgement is missing, adopt durable turns safely, and recover final sends across Telegram formatting and flood-wait failures. Thanks @100yenadmin, @Kyzcreig, @obviyus, and @jalehman.
**Startup and upgrade recovery:** run container migrations before Gateway readiness, **Fixed:** Recoverable legacy state from blocking startup, and enter control-plane-safe mode after repeated unclean boots instead of restart flapping. Fixes #98565 Thanks @sallyom, @jacobtomlinson, @bdjben, @obviyus, and @shakkernerd.
**Meta provider:** **Bundled `muse-spark-1.1` Responses API** is now supported as a provider with streaming, tool calls, encrypted reasoning replay, onboarding, exact-model live validation, and standalone npm/ClawHub distribution as `@openclaw/meta-provider`. Thanks @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, and @vincentkoc.
**Android chat agent selector:** switch the active agent directly from the live chat screen while keeping chat, Talk mode, and home canvas on the same canonical session. Thanks @bcperry and @joshavant.
**Gateway host status:** show the connected Gateway's host, network address, OS, runtime, uptime, CPU, memory, and disk details in Control UI Settings.
...and 27 more
🐛 Bug Fixes
**Codex runtime switching:** accept the bundled Codex runtime for both `codex/*` and `openai/*` model routes while keeping unsupported provider/runtime pairs rejected.
**LM Studio and xAI repaired tool calls:** emit the terminal tool-call event before completion when promoting serialized plain-text tool calls, preserving the complete streaming lifecycle for consumers.
**Browser actions on Node 24:** keep browser request cancellation bound to the client and response lifetime instead of Node 24.16+'s prematurely aborted body-stream signal, preventing valid POST actions from failing after JSON parsing. Thanks @obviyus and @vincentkoc.
**SecretRef model credentials:** keep resolved provider secrets behind process-local sentinels through auth storage, stream setup, SDK configuration, and managed local-provider probing, then inject plaintext only at the final network or provider-plugin boundary while retaining exact-value log redaction.
**Telegram token redaction:** redact bot tokens even when log transports split them across chunks, preventing fragmented credentials from escaping structured and streamed logs.
**Telegram durable turn adoption:** complete spooled updates only after durable agent adoption, detach adopted turns from the ingress reply fence, and suppress superseded replies so recovery cannot leak stale or duplicate responses. Thanks @obviyus and @vincentkoc.
**Task delivery recovery:** normalize legacy delivery statuses while restoring task records so older queued work becomes runnable instead of aborting registry recovery. Fixes #103168 Thanks @bek91, @theo674, and @obviyus.
**Diagnostics provider evidence:** emit one deduplicated `provider.request` timeline event for every completed or failed model call, so opt-in timelines can prove real provider traffic without double-counting terminal paths. Fixes #103063.
**Lean local model shell access:** keep `exec` directly visible beside the default structured Tool Search controls so coding-tuned local models can use their shell fallback instead of searching for missing domain tools. Thanks @vincentkoc and @maweibin.
**OAuth refresh contention diagnostics:** keep local lock paths out of user-facing refresh failures and **Fixed:** Duplicate failure prefixes while preserving structured provider and profile classification.
**New models and providers:** add Featherless, Claude Sonnet 5 and Mythos 5, Meta Muse Spark 1.1, and ClawRouter; make GPT-5.6 the new-setup default with `/think ultra` for Sol and Terra plus `max` for Luna, honor Z.AI `max`, and refresh model availability after OAuth renewal. Thanks @vincentkoc, @vortexopenclaw, @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, @steipete-oai, @bdjben, @obviyus, @sallyom, @anyech, @100yenadmin, and @fuller-stack-dev.
**Control UI and native macOS chat:** make sessions primary with a minimal searchable sidebar, compact context ring, reasoning-effort slider, cleaner dashboard chrome, and a native macOS session browser with model and thinking pickers, slash commands, transcript export, and context usage.
**Conversational onboarding:** Crestodian now runs a real agent-loop setup across the CLI, web install, and macOS app, with AI-guided provider setup, model-judged approvals bound to exact operations, masked credential prompts, and deterministic fallback when no model is available.
**Offline and spoken mobile chat:** iOS and Android now pre-paint bounded per-gateway session and transcript caches offline; Apple Watch gains full voice turns, and iOS can speak replies through configured Gateway TTS with on-device fallback.
**Session organization and generated titles:** generate concise session titles through utility-model routing and add Gateway-backed groups, unread state, rename, fork, archive, and delete controls, plus group management across web, iOS, and Android. Thanks @zhangguiping-xydt and @Juliangsm.
**Telegram and Codex continuity:** pair Codex through private `/login`, steer active runs with `/steer` and `/tell`, preserve messages when transcript acknowledgement is missing, adopt durable turns safely, and recover final sends across Telegram formatting and flood-wait failures. Thanks @100yenadmin, @Kyzcreig, @obviyus, and @jalehman.
**Startup and upgrade recovery:** run container migrations before Gateway readiness, **Fixed:** Recoverable legacy state from blocking startup, and enter control-plane-safe mode after repeated unclean boots instead of restart flapping. Fixes #98565 Thanks @sallyom, @jacobtomlinson, @bdjben, @obviyus, and @shakkernerd.
**Meta provider:** **Bundled `muse-spark-1.1` Responses API** is now supported as a provider with streaming, tool calls, encrypted reasoning replay, onboarding, exact-model live validation, and standalone npm/ClawHub distribution as `@openclaw/meta-provider`. Thanks @HamidShojanazeri, @davemorin, @Solvely-Colin, @jalehman, and @vincentkoc.
**Android chat agent selector:** switch the active agent directly from the live chat screen while keeping chat, Talk mode, and home canvas on the same canonical session. Thanks @bcperry and @joshavant.
**Gateway host status:** show the connected Gateway's host, network address, OS, runtime, uptime, CPU, memory, and disk details in Control UI Settings.
...and 27 more
🐛 Bug Fixes
**Codex runtime switching:** accept the bundled Codex runtime for both `codex/*` and `openai/*` model routes while keeping unsupported provider/runtime pairs rejected.
**LM Studio and xAI repaired tool calls:** emit the terminal tool-call event before completion when promoting serialized plain-text tool calls, preserving the complete streaming lifecycle for consumers.
**Browser actions on Node 24:** keep browser request cancellation bound to the client and response lifetime instead of Node 24.16+'s prematurely aborted body-stream signal, preventing valid POST actions from failing after JSON parsing. Thanks @obviyus and @vincentkoc.
**SecretRef model credentials:** keep resolved provider secrets behind process-local sentinels through auth storage, stream setup, SDK configuration, and managed local-provider probing, then inject plaintext only at the final network or provider-plugin boundary while retaining exact-value log redaction.
**Telegram token redaction:** redact bot tokens even when log transports split them across chunks, preventing fragmented credentials from escaping structured and streamed logs.
**Telegram durable turn adoption:** complete spooled updates only after durable agent adoption, detach adopted turns from the ingress reply fence, and suppress superseded replies so recovery cannot leak stale or duplicate responses. Thanks @obviyus and @vincentkoc.
**Task delivery recovery:** normalize legacy delivery statuses while restoring task records so older queued work becomes runnable instead of aborting registry recovery. Fixes #103168 Thanks @bek91, @theo674, and @obviyus.
**Diagnostics provider evidence:** emit one deduplicated `provider.request` timeline event for every completed or failed model call, so opt-in timelines can prove real provider traffic without double-counting terminal paths. Fixes #103063.
**Lean local model shell access:** keep `exec` directly visible beside the default structured Tool Search controls so coding-tuned local models can use their shell fallback instead of searching for missing domain tools. Thanks @vincentkoc and @maweibin.
**OAuth refresh contention diagnostics:** keep local lock paths out of user-facing refresh failures and **Fixed:** Duplicate failure prefixes while preserving structured provider and profile classification.
Fixes newer Google Chat direct messages sometimes being treated like group conversations, so they reach the correct one-to-one chat while Space and group-chat messages keep their existing routing. #58993 Thanks @starhappysh, @vincentkoc.
Feishu voice replies from OpenClaw now show their duration in the chat bubble, so recipients can see how long the audio is before playing it. #89172 Related #53798. Thanks @areslp, @fxz26284407, @kinrocw.
Discord and Telegram replies and mirrored chat history stay tied to the intended conversation more consistently, including across repeated Telegram replies and session changes. #89911.
Background image, video, and music results now return to the chat that requested them when the task starts without a full conversation target, instead of appearing to fail after creation or being sent to the wrong peer as the session moves. #89949 Related #86034. Thanks @tianxiaochannel-oss88, @wangwllu.
Telegram answers now stay attached to the user's current question when they quote an earlier bot message, while quotes of other people's messages still reply to the selected quote. #90475.
QQBot group admins can choose how broadly slash commands are available, and private-only commands now direct users to a private chat instead of being exposed or silently ignored in groups. #92154.
Heartbeat checks using reasoning-capable models now show the assistant's intended reply instead of exposing internal reasoning in Telegram, WhatsApp, and other channels, while opt-in Thinking messages still work. #92356 Related #92260. Thanks @jmpei, @tangtaizong666, @vincentkoc.
Telegram progress-mode chats now clear an old progress bubble before newer tool output or artifacts appear, keeping the conversation in a clean, readable order. #93002 Related #90753. Thanks @shadow-enthusiast, @zhangguiping-xydt.
IMessage command-and-link messages now stay together as one OpenClaw turn when delayed link previews arrive, while unrelated quick messages remain separate for users who enabled same-sender DM coalescing. #93143.
Successful Discord replies sent through the message tool no longer trigger a misleading failure warning in affected `message_tool_only` source-channel turns. #94072 Related #93875. Thanks @chenyangjun-xy, @hoyanhan, @vincentkoc.
**More capable channel control:** Slack relay mode, native Mattermost `/oc_queue`, and per-DM model overrides make channel operations easier to automate and tune. Thanks @sjf-oa, @amknight, @xydigit-zt, @thomaszta, and @gandalf-at-lerian.
**Richer operator workflows:** `openclaw agent --message-file` and the RAFT CLI wake bridge add practical file-driven and remote wake-up paths. Thanks @ooiuuii and @vincentkoc.
**Safer plugin distribution:** additional official plugins are externalized cleanly, with bundled plugin icon metadata available to installed clients. Thanks @vincentkoc and @Patrick-Erichsen.
**Stronger mobile operations:** Android settings detail panels **Improved:** Configuration visibility and control on mobile.
**More reliable agent turns:** Codex partial deltas, harness activation, and long-context prompt-cache stability reduce lost progress and inconsistent runs. Thanks @agonza1 and @vincentkoc.
**Gateway and plugin tooling:** channel identity hook context and per-agent usage-cost reporting give integrations and operators more precise routing and accounting. Thanks @lanzhi-lee, @vincentkoc, and @ly-wang19.
**Provider and model coverage:** catalog parsing, reasoning controls, provider model resolution, and encrypted reasoning support now handle more live provider variants. Thanks @ZengWen-DT, @vincentkoc, @Marvinthebored, @Darren2030, @daniel-alejandro-t, @parveshsaini, @geraint0923, @fuller-stack-dev, and @jason-allen-oneal.
🐛 Bug Fixes
**Channel delivery:** Telegram progress rendering, webhook lifecycle, reaction directives, duplicate mirror writes, queued update draining, and WhatsApp durable reply targets are now more reliable. Thanks @amknight, @snowzlmbot, @zhangguiping-xydt, @shadow-enthusiast, @xialonglee, @travellingsoldier85, @obviyus, @hugenshen, @Cuttingwater, @heichaowo, @LiuwqGit, @freidrich-goldenflow, @mcaxtr, and @vincentkoc.
**WhatsApp and message identity:** native quotes, Baileys group reliability, and approval reactions across JID drift now preserve the intended conversation context. Thanks @mcaxtr, @xialonglee, and @octopuslabs-fl.
**Gateway and session safety:** stuck release claims, draining-state reporting, remote probe timeouts, malformed paired access lists, and non-delivery session identity are handled without silent routing loss. Thanks @mikasa0818, @kriegerbangerz-ship-it, @markoub, @vincentkoc, @maxschachere, @mushuiyu886, @gozzbb2, @wangmiao0668000666, @ly-wang19, @EmilioNicolas, @yetval, and @hellocli.
**Agent and fallback behavior:** aborted runs **Fixed:** Cleanly, provider response bodies stay bounded, Claude CLI credit failures continue through fallback, and Codex usage-limit responses classify correctly. Thanks @szsip239, @vincentkoc, @Alix-007, @mikasa0818, @sallyom, @riazrahaman, and @jason-allen-oneal.
**Provider and model edge cases:** OpenRouter IDs, Ollama discovery and embeddings, Gemini freshness, and model-catalog prefixes now **Fixed:** Against the right runtime metadata. Thanks @Darren2030, @daniel-alejandro-t, @mushuiyu886, @jason-allen-oneal, @Sunjae-k, @parveshsaini, @vincentkoc, and @shakkernerd.
**Configuration and UI guardrails:** non-interactive configure fails closed, TLS paths reject empty values, memory artifacts are sanitized, and the UI uses the patched DOMPurify release. Thanks @ruomuxydt, @NianJiuZst, @miorbnli, @vincentkoc, @SweetSophia, and @YB0y.
**Cron and delivery validation:** no-config delivery checks, thread-aware dedupe, and pending recurring runs retain their intended destinations. Thanks @vincentkoc and @yetval.
**PR #95406** test(qa): make release scorecard categories explicit.
**PR #94700** test: fold HTTP API script proof into QA Lab.
**More capable channel control:** Slack relay mode, native Mattermost `/oc_queue`, and per-DM model overrides make channel operations easier to automate and tune. Thanks @sjf-oa, @amknight, @xydigit-zt, @thomaszta, and @gandalf-at-lerian.
**Richer operator workflows:** `openclaw agent --message-file` and the RAFT CLI wake bridge add practical file-driven and remote wake-up paths. Thanks @ooiuuii and @vincentkoc.
**Safer plugin distribution:** additional official plugins are externalized cleanly, with bundled plugin icon metadata available to installed clients. Thanks @vincentkoc and @Patrick-Erichsen.
**Stronger mobile operations:** Android settings detail panels **Improved:** Configuration visibility and control on mobile.
**More reliable agent turns:** Codex partial deltas, harness activation, and long-context prompt-cache stability reduce lost progress and inconsistent runs. Thanks @agonza1 and @vincentkoc.
**Gateway and plugin tooling:** channel identity hook context and per-agent usage-cost reporting give integrations and operators more precise routing and accounting. Thanks @lanzhi-lee, @vincentkoc, and @ly-wang19.
**Provider and model coverage:** catalog parsing, reasoning controls, provider model resolution, and encrypted reasoning support now handle more live provider variants. Thanks @ZengWen-DT, @vincentkoc, @Marvinthebored, @Darren2030, @daniel-alejandro-t, @parveshsaini, @geraint0923, @fuller-stack-dev, and @jason-allen-oneal.
🐛 Bug Fixes
**Channel delivery:** Telegram progress rendering, webhook lifecycle, reaction directives, duplicate mirror writes, queued update draining, and WhatsApp durable reply targets are now more reliable. Thanks @amknight, @snowzlmbot, @zhangguiping-xydt, @shadow-enthusiast, @xialonglee, @travellingsoldier85, @obviyus, @hugenshen, @Cuttingwater, @heichaowo, @LiuwqGit, @freidrich-goldenflow, @mcaxtr, and @vincentkoc.
**WhatsApp and message identity:** native quotes, Baileys group reliability, and approval reactions across JID drift now preserve the intended conversation context. Thanks @mcaxtr, @xialonglee, and @octopuslabs-fl.
**Gateway and session safety:** stuck release claims, draining-state reporting, remote probe timeouts, malformed paired access lists, and non-delivery session identity are handled without silent routing loss. Thanks @mikasa0818, @kriegerbangerz-ship-it, @markoub, @vincentkoc, @maxschachere, @mushuiyu886, @gozzbb2, @wangmiao0668000666, @ly-wang19, @EmilioNicolas, and @yetval.
**Agent and fallback behavior:** aborted runs **Fixed:** Cleanly, provider response bodies stay bounded, Claude CLI credit failures continue through fallback, and Codex usage-limit responses classify correctly. Thanks @szsip239, @vincentkoc, @Alix-007, @mikasa0818, @sallyom, @riazrahaman, and @jason-allen-oneal.
**Provider and model edge cases:** OpenRouter IDs, Ollama discovery and embeddings, Gemini freshness, and model-catalog prefixes now **Fixed:** Against the right runtime metadata. Thanks @Darren2030, @daniel-alejandro-t, @mushuiyu886, @jason-allen-oneal, @Sunjae-k, @parveshsaini, @vincentkoc, and @shakkernerd.
**Configuration and UI guardrails:** non-interactive configure fails closed, TLS paths reject empty values, memory artifacts are sanitized, and the UI uses the patched DOMPurify release. Thanks @ruomuxydt, @NianJiuZst, @miorbnli, @vincentkoc, @SweetSophia, and @YB0y.
**Cron and delivery validation:** no-config delivery checks, thread-aware dedupe, and pending recurring runs retain their intended destinations. Thanks @vincentkoc and @yetval.
**PR #95406** test(qa): make release scorecard categories explicit.
**PR #94700** test: fold HTTP API script proof into QA Lab.
Adds `/fast auto` so short conversational calls can start quickly, while longer or fallback work returns to normal mode with the effective state still visible. PR #85104, Issue #85087. Thanks @alexph-dev and @vincentkoc.
Shows the effective automatic fast-mode state in status instead of reducing it to on/off, and avoids carrying a cleared Codex service-tier choice into later runs. 8845f2f.
Keeps automatic fast-mode timing consistent when a turn switches to a fallback model. 075091d.
Keeps the original fast-mode timing and progress behavior when a live model switch retries a turn. d1e190f.
Keeps automatic fast-mode progress and reset behavior distinct from explicit fast mode after a run switches modes. 20aec98.
Shows the effective fast-mode value in connected-agent sessions instead of the configured value, so status reflects what the session is actually using. 9509aa0.
Keeps the effective automatic fast-mode setting visible through fallback transitions in connected-agent sessions. 7f5423c.
Keeps automatic fast-mode timing and progress consistent when reply and scheduled-agent runs retry or switch models. 6c29f88.
Keeps fast-mode cleanup and status consistent when a run switches between fallback models. c4694f8.
Shows the automatic fast-mode reset only when fallback work is finished, so status messages match the end of the transition. f4d93c8.
**Automatic fast mode for talks:** OpenClaw can Enabled fast mode for short conversational turns, then return to normal mode for longer runs with bounded fallback and delivery behavior. Thanks @alexph-dev and @vincentkoc.
**More reliable model routing:** Zai model synthesis, GLM overload failover, and native reasoning-level selection now follow the active model catalog more consistently. Thanks @Pandah97, @chrysb, @0xghost42, @zhengli0922, @openperf, @civiltox, and @BorClaw.
**Safer session and channel state:** channel switches reset stale origin fields, and cron delivery awareness stays attached to the target session. Thanks @ZengWen-DT, @jalehman, @gorkem2020, and @scotthuang.
**Trusted policies survive hook composition:** composed hook registries keep the trusted tool policies required by approval-sensitive flows.
**Agent and channel runtime:** fast-mode state now survives retries, fallback transitions, progress events, and embedded/CLI/ACP normalization; session and channel routing retain the current target and delivery context. Thanks @alexph-dev, @vincentkoc, @scotthuang, @ZengWen-DT, @jalehman, and @gorkem2020.
**Provider behavior:** model catalogs now supply the correct Zai base URL, overload classification, and native reasoning controls for live-discovered models. Thanks @Pandah97, @chrysb, @0xghost42, @zhengli0922, @openperf, @civiltox, and @BorClaw.
🐛 Bug Fixes
**Fast-mode and policy correctness:** fallback cutoffs and reset notices are bounded, repeated progress events remain visible, Codex service-tier state is normalized, and trusted policies are not lost when hook registries are composed. Thanks @alexph-dev, @vincentkoc, and @jesse-merhi.
**Model and delivery edge cases:** Zai and GLM failover paths use the right runtime metadata, while stale channel-origin state no longer leaks across session changes. Thanks @Pandah97, @chrysb, @0xghost42, @zhengli0922, @ZengWen-DT, @jalehman, and @gorkem2020.
**PR #86627** Keep core doctor health in contribution order.
**PR #93580** fix: preserve cron delivery awareness for target sessions. Thanks @scotthuang and @jalehman.
**PR #95328** fix(sessions): reset stale per-channel origin fields on channel switch. Related #95325. Thanks @ZengWen-DT and @jalehman and @gorkem2020.
**PR #94461** fix(zai): fall back to manifest baseUrl for synthesized GLM-5 models. Related #94269. Thanks @Pandah97 and @chrysb.
**PR #93241** fix(agents): classify Zhipu GLM overload as overloaded for failover. Related #93211. Thanks @0xghost42 and @zhengli0922.
**PR #94067** fix(channels): **Fixed:** Native /think menu levels via runtime catalog for live-discovered models. Related #93835. Thanks @openperf and @civiltox.
**More reliable agent turns and session state:** OpenClaw now preserves pending subagent completion announcements, keeps chat history transcripts non-empty, maintains media index alignment, restarts dormant follow-up drains, and resolves compaction model aliases consistently. Thanks @sallyom, @oiGaDio, @Hidetsugu55, @Nas01010101, @SFVVC, @Pick-cat, and @vincentkoc.
**Stronger Codex and approval flows:** Codex app-server SecretRefs, thread context, bounded turn text, routed approval context, and typed SDK approval/session helpers now work together more predictably. Thanks @VACInc, @kevinlin-openai, @kevinslin, @Nas01010101, @849261680, @choury, and @vincentkoc.
**Richer channel delivery:** Telegram, Discord, and Slack now preserve richer progress/reasoning/thread output, handle structured send errors, support Slack shortcuts, and record canonical sent threads more reliably. Thanks @obviyus, @zhangqueping, @jairrab, @ZOOWH, @parveshsaini, @vincentkoc, @Marvinthebored, @chrisbaker2000, @KennanHoa, @bek91, and @rockyloveswine.
**Safer release and network boundaries:** SSH tunnel preflight is loopback-scoped, device-backed node pairings are removed, volatile SQLite state is surfaced by doctor, and legacy Codex routes are repaired instead of silently persisting stale state. Thanks @wangwllu, @Alix-007, @manju-rn, @vincentkoc, @TurboTheTurtle, and @Sleepyarno.
**Useful new CLI and status workflows:** Rename sessions from chat, compact sessions explicitly, show session duration, preserve command progress detail, and preview message sends/polls with dry-run output. Thanks @BSG2000, @Alix-007, @sallyom, @redasadki, @marshall-gordfam, @vincentkoc, @lzyyzznl, and @YB0y.
**More capable mobile and desktop clients:** Android settings are grouped by intent, iOS notification state is cleaner, the Watch app uses the Xcode 27-compatible target layout, and macOS file inputs open through the native panel. Thanks @Tosko4, @zats, @joshavant, @bbblending, @DINGDANGMAOUP, and @vincentkoc.
**Broader plugin and skill coverage:** Zalo is available as an external channel entry, Trello skills declare their curl dependency, stale managed skill links are retargeted, and tool discovery no longer clears active providers. Thanks @ken-kuro, @liuhao1024, @berkgungor, @stevenepalmer, @shakkernerd, @medns, and @vincentkoc.
**Agent and provider behavior:** Codex turn limits, CLI-owned auth, provider-internal error wording, recurring cron backoff, explicit cron delivery targets, and isolated cron key requirements now fail or recover more clearly. Thanks @Nas01010101, @yu-xin-c, @snowzlmbot, @Alix-007, @jincheng-xydt, @sallyom, @davectr, @hxy91819, @nxmxbbd, and @vincentkoc.
**Channels and integrations:** WhatsApp retries the opening text chunk after media failure, Feishu avoids axios internals, Slack records inbound mentions and preserves buffered streams, and external Zalo/Slack shortcuts are wired through the current channel seams. Thanks @yetval, @sweetcornna, @davinci282828, @ZengWen-DT, @BryceMurray, @vincentkoc, @KennanHoa, @ken-kuro, and @chrisbaker2000.
**Skills and setup:** OnePassword auth no longer forces tmux when the desktop app is available, stale plugin skill symlinks are repaired, and Trello requirements match their examples. Thanks @koshaji, @tylerbittner, @stevenepalmer, @shakkernerd, @liuhao1024, @berkgungor, and @vincentkoc.
...and 2 more
🐛 Bug Fixes
**Reply and transcript correctness:** OpenClaw now keeps pending completions, non-empty histories, media fields, queued follow-ups, buffered Slack replies, and reasoning deliveries intact across retries and partial turns. Thanks @sallyom, @oiGaDio, @Hidetsugu55, @Nas01010101, @SFVVC, @vincentkoc, @KennanHoa, @Marvinthebored, @zerone0x, and @pearl-dot.
**Security and bounded input handling:** SSH tunnel checks stay on loopback, unsafe chat/tool/package/response lengths are rejected, device-backed pairings are removed, and stale abort markers no longer affect fresh chat events. Thanks @wangwllu, @vincentkoc, @Alix-007, @manju-rn, and @nxmxbbd.
**Telegram, WhatsApp, and Slack delivery:** Rich progress previews, structured Telegram errors, WhatsApp listener recovery, and canonical Slack thread/send behavior now survive the edge cases covered by the new release. Thanks @obviyus, @zhangqueping, @jairrab, @ZOOWH, @parveshsaini, @xialonglee, @octaivermatt, @bek91, and @rockyloveswine.
**Cron and queue safety:** Recurring error backoff honors configured floors, implicit isolated delivery requires an explicit target, and dormant follow-up drains restart instead of disappearing. Thanks @Alix-007, @nxmxbbd, @hxy91819, and @SFVVC.
**Provider, auth, and migration repair:** CLI-owned transports skip the wrong auth gate, compaction aliases **Fixed:** Canonically, legacy Codex routes are repaired, and tool discovery no longer clears active providers. Thanks @yu-xin-c, @Pick-cat, @TurboTheTurtle, @Sleepyarno, @medns, and @vincentkoc.
**SDK and release tooling:** Approval/session RPC params are typed more strictly, stale packed tarballs are ignored, and DMG output directories are created reliably.
**Richer Telegram delivery:** Telegram now sends rich HTML, preserves rich markdown and sticker paths, renders progress drafts and command output more faithfully, normalizes HTML tables safely, and keeps mentions and spooled handlers on the right delivery path. Thanks @obviyus, @vincentkoc, @goutamadwant, @kesslerio, @NianJiuZst, @SweetSophia, @Marvinthebored, @aaajiao, @zhangqueping, and @jairrab.
**More dependable agent recovery:** retries, terminal outcomes, usage after compaction, session history repair, and reply reconciliation now keep more interrupted or partial turns moving toward a visible final result. Thanks @ai-hpc, @lml2468, @fuller-stack-dev, @Hollychou924, @leno23, @de1tydev, @425072024, @wuwahe3, @drvoss, @yetval, @sandieman2, and @vincentkoc.
**A stronger Codex integration:** Codex gains automatic plugin approvals, GPT-5.3 Spark OAuth routing, remote-node `exec` as a dynamic tool, and more reliable app-server teardown and terminal outcomes. Thanks @kevinslin, @VACInc, @vincentkoc, @JPKay-AI, and @aliahnaf2013-max.
**Standalone official provider plugins:** external provider packages are now first-class npm releases, externally installed channel plugins load at Gateway startup, and StepFun is available from npm and ClawHub. Thanks @sunlit-deng, @cxdnicole, and @vincentkoc.
**More capable web and native clients:** the Control UI adds a session workspace rail and extension health, iOS adds Watch controls, and Android shows chat context. Thanks @Solvely-Colin, @jalehman, @joshavant, and @Tosko4.
**More useful search and skills:** Codex Hosted Search is available, key-free search providers remain deliberate opt-ins, and ClawHub skill installs retain verified source provenance. Thanks @fuller-stack-dev, @davemorin, @momothemage, @nmccready-tars, and @vincentkoc.
**🔒 Providers and auth:** Add Codex Hosted Search, improve Gemini CLI OAuth behind proxies, and keep external provider onboarding on current choices and package metadata. Thanks @fuller-stack-dev, @yetval, @EvetteYoung, and @vincentkoc.
**Plugins and installs:** Externalized official providers publish as independent npm packages, Gateway discovers installed channel plugins at startup, and StepFun installs from npm or ClawHub. Thanks @sunlit-deng, @cxdnicole, and @vincentkoc.
**Dashboard and mobile:** Add a session workspace rail, plugin health in status, compact cron lists, and iOS Watch controls. Thanks @Solvely-Colin, @jalehman, @yu-xin-c, @centralpc, @joshavant, and @vincentkoc.
Codex, observability, and skills: add automatic plugin approvals and SecretRefs, preserve ClawHub skill provenance, add OpenTelemetry log export, and expose remote-node execution to Codex when a node is connected. Thanks @kevinslin, @kevinlin-openai, @momothemage, @nmccready-tars, @jesse-merhi, @vincentkoc, and @JPKay-AI.
...and 1 more
🐛 Bug Fixes
**🔒 Security and privacy:** Redact secrets from debug/config output, block internal HTTP session overrides, audit open-DM tool exposure, and retain plugin write ownership checks. Thanks @Alix-007, @jason-allen-oneal, @coygeek, @RichardCao, @yu-xin-c, @cjg20ss, @eleqtrizit, and @vincentkoc.
**Agent and session runtime:** Retry thinking-only and empty post-tool turns, prevent duplicate hook execution, preserve pending subagent delivery, preserve fresh usage through compaction, and repair partial JSON/history artifacts. Thanks @ai-hpc, @lml2468, @fuller-stack-dev, @zenglingbiao, @dertbv, @Hollychou924, @leno23, @de1tydev, @425072024, @wuwahe3, @drvoss, @vincentkoc, @sallyom, @oiGaDio, @Hidetsugu55, and @Nas01010101.
**Channels and replies:** Fix Telegram rich delivery, table rendering, action-error handling, and ingress recovery; preserve command progress detail across channel adapters; retain WhatsApp opening text after a media failure; keep Mattermost thread replies intact; and harden Discord action handling. Thanks @obviyus, @NianJiuZst, @mcaxtr, @rushindrasinha, @amknight, @lzyyzznl, @darealgege, @vincentkoc, @zhangqueping, @jairrab, @ZOOWH, @parveshsaini, and @yetval.
**Storage and migrations:** Avoid SQLite WAL on network filesystems, clean reindex artifacts, keep setup state out of workspace dot-directories, and import default-agent auth profiles into SQLite. Thanks @vincentkoc, @ZengWen-DT, @Zeng-wen, @potterdigital, @Alix-007, @Pick-cat, @sallyom, @1qh, and @Tazio7.
**Provider and model behavior:** Fix Gemini CLI proxy OAuth, restore Codex Spark OAuth routing, correct Bedrock embedding model IDs, and preserve configured defaults in embedded runs. Thanks @yetval, @EvetteYoung, @VACInc, @LiuwqGit, @aleck31, @zenglingbiao, @danielgerlag, and @vincentkoc.
CLI, TUI, and apps: accept global flags after subcommands, keep terminal output and activity indicators visible, preserve CJK IME composition, and refresh stale UI state. Thanks @ooiuuii, @Alix-007, @ZengWen-DT, @Zeng-wen, @AlethiaQuizForge, @Zhaoqj2016, @liuhao1024, @BrianClaw1955, @vincentkoc, and @NicoBoom13.
**Operations and updates:** Harden official plugin recovery, restart managed Gateways after failed update handoff, keep safe cron delivery defaults, avoid Node-specific npm prefixes, and keep package validation paths reliable. Thanks @vincentkoc, @yetval, @ofan, @yaanfpv, @jincheng-xydt, @sallyom, @davectr, and @nxmxbbd.
**PR #90463** refactor: add session accessor seam with gateway consumer.
**PR #88656** Drop reasoning-only length turns from replay.
**Richer Telegram delivery:** Telegram now sends rich HTML, preserves rich markdown and sticker paths, renders progress drafts and command output more faithfully, and keeps mentions and spooled handlers on the right delivery path. Thanks @obviyus, @vincentkoc, @goutamadwant, @kesslerio, @NianJiuZst, @SweetSophia, @Marvinthebored, and @aaajiao.
**More dependable agent recovery:** retries, terminal outcomes, usage after compaction, session history repair, and reply reconciliation now keep more interrupted or partial turns moving toward a visible final result. Thanks @ai-hpc, @lml2468, @fuller-stack-dev, @Hollychou924, @leno23, @de1tydev, @425072024, @wuwahe3, @drvoss, @yetval, @sandieman2, and @vincentkoc.
**A stronger Codex integration:** Codex gains automatic plugin approvals, GPT-5.3 Spark OAuth routing, remote-node `exec` as a dynamic tool, and more reliable app-server teardown and terminal outcomes. Thanks @kevinslin, @VACInc, @vincentkoc, @JPKay-AI, and @aliahnaf2013-max.
**Standalone official provider plugins:** external provider packages are now first-class npm releases, externally installed channel plugins load at Gateway startup, and StepFun is intentionally npm-only because its ClawHub package name is unavailable. Thanks @sunlit-deng, @cxdnicole, and @vincentkoc.
**More capable web and native clients:** the Control UI adds a session workspace rail and extension health, iOS adds Watch controls, and Android shows chat context. Thanks @Solvely-Colin, @jalehman, @joshavant, and @Tosko4.
**More useful search and skills:** Codex Hosted Search is available, key-free search providers remain deliberate opt-ins, and ClawHub skill installs retain verified source provenance. Thanks @fuller-stack-dev, @davemorin, @momothemage, @nmccready-tars, and @vincentkoc.
**🔒 Providers and auth:** Add Codex Hosted Search, improve Gemini CLI OAuth behind proxies, and keep external provider onboarding on current choices and package metadata. Thanks @fuller-stack-dev, @yetval, @EvetteYoung, and @vincentkoc.
**Plugins and installs:** Externalized official providers publish as independent npm packages, Gateway discovers installed channel plugins at startup, and StepFun installs exclusively from npm. Thanks @sunlit-deng, @cxdnicole, and @vincentkoc.
**Dashboard and mobile:** Add a session workspace rail, plugin health in status, compact cron lists, and iOS Watch controls. Thanks @Solvely-Colin, @jalehman, @yu-xin-c, @centralpc, @joshavant, and @vincentkoc.
**Codex and skills:** Add automatic plugin approvals, preserve ClawHub skill provenance, and expose remote-node execution to Codex when a node is connected. Thanks @kevinslin, @momothemage, @nmccready-tars, @vincentkoc, and @JPKay-AI.
...and 1 more
🐛 Bug Fixes
**🔒 Security and privacy:** Redact secrets from debug/config output, block internal HTTP session overrides, audit open-DM tool exposure, and retain plugin write ownership checks. Thanks @Alix-007, @jason-allen-oneal, @coygeek, @RichardCao, @yu-xin-c, @cjg20ss, @eleqtrizit, and @vincentkoc.
**Agent and session runtime:** Retry thinking-only and empty post-tool turns, prevent duplicate hook execution, preserve fresh usage through compaction, and repair partial JSON/history artifacts. Thanks @ai-hpc, @lml2468, @fuller-stack-dev, @zenglingbiao, @dertbv, @Hollychou924, @leno23, @de1tydev, @425072024, @wuwahe3, @drvoss, and @vincentkoc.
**Channels and replies:** Fix Telegram rich delivery and ingress recovery, preserve WhatsApp auth and media error reporting, keep Mattermost thread replies intact, and harden Discord action handling. Thanks @obviyus, @NianJiuZst, @mcaxtr, @rushindrasinha, @amknight, @lzyyzznl, @darealgege, and @vincentkoc.
**Storage and migrations:** Avoid SQLite WAL on network filesystems, clean reindex artifacts, keep setup state out of workspace dot-directories, and import default-agent auth profiles into SQLite. Thanks @vincentkoc, @ZengWen-DT, @Zeng-wen, @potterdigital, @Alix-007, @Pick-cat, @sallyom, @1qh, and @Tazio7.
**Provider and model behavior:** Fix Gemini CLI proxy OAuth, restore Codex Spark OAuth routing, correct Bedrock embedding model IDs, and preserve configured defaults in embedded runs. Thanks @yetval, @EvetteYoung, @VACInc, @LiuwqGit, @aleck31, @zenglingbiao, @danielgerlag, and @vincentkoc.
CLI, TUI, and apps: accept global flags after subcommands, keep terminal output and activity indicators visible, preserve CJK IME composition, and refresh stale UI state. Thanks @ooiuuii, @Alix-007, @ZengWen-DT, @Zeng-wen, @AlethiaQuizForge, @Zhaoqj2016, @liuhao1024, @BrianClaw1955, @vincentkoc, and @NicoBoom13.
**Operations and updates:** Harden official plugin recovery, restart managed Gateways after failed update handoff, avoid Node-specific npm prefixes, and keep package validation paths reliable. Thanks @vincentkoc, @yetval, @ofan, and @yaanfpv.
**PR #90463** refactor: add session accessor seam with gateway consumer.
**PR #88656** Drop reasoning-only length turns from replay.
**Richer channel delivery:** Telegram and WhatsApp are less brittle: Telegram renders structured text with tables, lists, expandable blockquotes, preserved intentional line breaks, and CLI-backed replies, while WhatsApp now honors configured ACP bindings. Thanks @obviyus, @vincentkoc, @jzakirov, @spacegeologist, @TurboTheTurtle, @mcaxtr, @myrzka, and @dmorn.
**More reliable agent runs:** account-scoped DM sends, generated media completions, auto-reply message-tool final replies, reset archive fallback reads, restart shutdown aborts, yielded subagent pauses, and session identity prompts all stay on the correct recovery path. Thanks @yetval, @TurboTheTurtle, @masatohoshino, @CadanHu, @vincentkoc, @ooiuuii, @openperf, @zhangguiping-xydt, @QQSHI13, @kumaxs, and @aleps001.
**Safer model routing:** new GLM-5.2 and Claude Haiku 4.5 catalog support arrives with normalized provider IDs, managed SecretRef auth, bounded model browsing, and safer OpenAI/Anthropic tool-schema recovery. Thanks @arkyu2077, @liuhao1024, @lijenhsin, @rohitjavvadi, @samson910022, @maaron34, @syfvb, and @samson1357924.
**Useful usage footers:** `/usage` and reply payload hooks now have a native full footer renderer, default template, fixed-decimal formatting, credential-aware limits, **Improved:** Partial-count handling, and warnings for broken templates instead of silent bad output.
**Predictable web search defaults:** key-free providers such as Parallel Free, DuckDuckGo, Ollama, and Codex Hosted Search remain explicit opt-ins rather than surprising automatic fallbacks. Thanks @davemorin and @vincentkoc.
**Calmer UI and mobile sessions:** workspace files start collapsed, WebChat backscroll survives streaming, the desktop session picker remains interactive, reset arguments survive dispatch, and iOS reconnects stale foreground Gateways. Thanks @shakkernerd, @TurboTheTurtle, @NianJiuZst, @zhouhe-xydt, @Solvely-Colin, @MaBeitian, @vincentkoc, @Chang2020618, and @DrtyMorty.
**Resilient memory and state:** oversized OpenAI embedding batches split before 431s, QMD search stays available in transient mode, SQLite avoids WAL on NFS volumes, and full reindexes preserve rollback/cache recovery. Thanks @mushuiyu886, @BrettHamlin, @zhbcher, @TurboTheTurtle, @Takhoffman, @849261680, @TSHOGX, @vincentkoc, and @AFabyTWE.
**Providers/models:** Add GLM-5.2 support and Claude Haiku 4.5 catalog entries while keeping provider-qualified model IDs normalized across OpenRouter and Google Vertex paths. Thanks @arkyu2077, @liuhao1024, @bymle, @maaron34, @lijenhsin, @davemorin, and @vincentkoc.
**UI:** Keep key-free providers such as Parallel Free, DuckDuckGo, Ollama, and Codex Hosted Search as explicit opt-ins instead of selecting them automatically when no API-backed provider is configured. Thanks @davemorin and @vincentkoc.
**Channel plugins:** Ship Telegram rich-message delivery and WhatsApp ACP binding support, including preserved intentional line breaks, rich prompt handoff to CLI backends, and transport fixtures for richer drafts. Thanks @obviyus, @TurboTheTurtle, @vincentkoc, @mcaxtr, and @dmorn.
...and 2 more
🐛 Bug Fixes
**Channels and delivery:** Preserve account-scoped DM channel send policy, intentional rich-message line breaks in Telegram and status output, rich Telegram final replies, rich Telegram tables and lists, Telegram thread-create CLI remapping, Feishu dynamic-agent routes after persisted binding reuse, Slack outbound `message_sent` hooks, contributed message-tool schema optionality, same-channel generated media completions, and channel chunking around surrogate pairs and Infinity limits. Thanks @yetval, @obviyus, @spacegeologist, @rishitamrakar, @liuhao1024, @lundog, @TurboTheTurtle, @yhterrance, @vincentkoc, @myrzka, @cwlong163-afk, @kumaxs, @shakkernerd, and @RewardsPal.
**CLI:** Use the selected OpenClaw OAuth/API-key auth profile in an isolated Gemini CLI runtime home, preventing ambient Google machine credentials from overriding the chosen profile. Thanks @jason-allen-oneal and @shakkernerd.
**Discord:** Give generated auto-thread titles a 60-second timeout and 4,096-token reasoning-model output budget, clamped to the selected model output cap.
Agent, cron, and Gateway runtime: mark active main sessions before restart shutdown aborts, pause yielded subagent runs whose terminal also signals abort, clamp trusted subagent thinking overrides through provider/model fallback, preserve yielded media completions, deliver channel message-tool final replies through auto-reply while hiding internal delivery hints, restore reset archive fallback reads when active async transcripts are missing, de-duplicate main-session heartbeat events, expose session identity in runtime prompts, reject unknown OpenAI agent selectors, keep generated media completions, slash-command block replies, and trajectory export commands in WebChat, and require admin privileges for HTTP session/model override surfaces. Thanks @ooiuuii, @openperf, @IWhatsskill, @masatohoshino, @CadanHu, @ZengWen-DT, @zhangguiping-xydt, @TurboTheTurtle, @oiGaDio, @aleps001, @vincentkoc, @GSL-R, @QQSHI13, @ryanhelms, @kumaxs, @steipete-oai, @hxy91819, @davemorin, and @nailujac.
**Providers and model replay:** Preserve storeless OpenAI Responses replay compatibility, recover invalid OpenAI reasoning signatures and genericized Anthropic thinking-signature replay errors, route OAuth image defaults through Codex for eligible OpenAI profiles, avoid eager tool streaming for Claude 4.5 in Copilot, quarantine unreadable and post-hook OpenAI/Anthropic-family tool schemas without broadening allowed tool choices, deliver explicit thinking-off requests to LM Studio binary-thinking models, honor profile auth for SecretRef model entries, bound model browsing, strip provider prefixes where runtimes need bare IDs, and surface nested embedding fetch failures. Thanks @snowzlm, @mmyzwl, @CarlCapital, @bek91, @Kailigithub, @vincentkoc, @rohitjavvadi, @samson910022, @nxmxbbd, @liuhao1024, @bymle, @mushuiyu886, @finchinslc, @syfvb, @lijenhsin, @crsnpalmer-art, @samson1357924, @shakkernerd, and @mlaihk.
Memory, state, diagnostics, and config: split header-too-large embedding batches, keep QMD memory search enabled in transient mode, **Fixed:** SQLite WAL on NFS volumes, preserve recovery scheduling outside stuck-session warning backoff, preserve full-reindex rollback/cache recovery, and treat raw Memory Wiki source pages as source evidence. Thanks @mushuiyu886, @TurboTheTurtle, @849261680, @gnanam1990, @TSHOGX, @vincentkoc, @arlen8411, @BrettHamlin, @zhbcher, @Takhoffman, @AFabyTWE, @davemorin, and @zhuyankarl.
**Plugins and updates:** Repair missing required platform packages during managed plugin installs and updates, including omitted Codex platform binaries.
**Dependencies:** Update Hono to 4.12.25 so published OpenClaw and ACPX packages use the patched runtime.
**Updates:** Avoid a false downgrade prompt when the latest tag cannot resolve. Thanks @Andy312432 and @vincentkoc.
**Richer channel delivery:** Telegram and WhatsApp are less brittle: Telegram renders structured text with tables, lists, expandable blockquotes, preserved intentional line breaks, and CLI-backed replies, while WhatsApp now honors configured ACP bindings. Thanks @obviyus, @vincentkoc, @jzakirov, @spacegeologist, @TurboTheTurtle, @mcaxtr, @myrzka, and @dmorn.
**More reliable agent runs:** account-scoped DM sends, generated media completions, auto-reply message-tool final replies, reset archive fallback reads, restart shutdown aborts, yielded subagent pauses, and session identity prompts all stay on the correct recovery path. Thanks @yetval, @TurboTheTurtle, @masatohoshino, @CadanHu, @vincentkoc, @ooiuuii, @openperf, @zhangguiping-xydt, @QQSHI13, @kumaxs, and @aleps001.
**Safer model routing:** new GLM-5.2 and Claude Haiku 4.5 catalog support arrives with normalized provider IDs, managed SecretRef auth, bounded model browsing, and safer OpenAI/Anthropic tool-schema recovery. Thanks @arkyu2077, @liuhao1024, @lijenhsin, @rohitjavvadi, @samson910022, @maaron34, @syfvb, and @samson1357924.
**Useful usage footers:** `/usage` and reply payload hooks now have a native full footer renderer, default template, fixed-decimal formatting, credential-aware limits, **Improved:** Partial-count handling, and warnings for broken templates instead of silent bad output.
**Predictable web search defaults:** key-free providers such as Parallel Free, DuckDuckGo, Ollama, and Codex Hosted Search remain explicit opt-ins rather than surprising automatic fallbacks. Thanks @davemorin and @vincentkoc.
**Calmer UI and mobile sessions:** workspace files start collapsed, WebChat backscroll survives streaming, the desktop session picker remains interactive, reset arguments survive dispatch, and iOS reconnects stale foreground Gateways. Thanks @shakkernerd, @TurboTheTurtle, @NianJiuZst, @zhouhe-xydt, @Solvely-Colin, @MaBeitian, @vincentkoc, @Chang2020618, and @DrtyMorty.
**Resilient memory and state:** oversized OpenAI embedding batches split before 431s, QMD search stays available in transient mode, SQLite avoids WAL on NFS volumes, and full reindexes preserve rollback/cache recovery. Thanks @mushuiyu886, @BrettHamlin, @zhbcher, @TurboTheTurtle, @Takhoffman, @849261680, @TSHOGX, @vincentkoc, and @AFabyTWE.
**Providers/models:** Add GLM-5.2 support and Claude Haiku 4.5 catalog entries while keeping provider-qualified model IDs normalized across OpenRouter and Google Vertex paths. Thanks @arkyu2077, @liuhao1024, @bymle, @maaron34, @lijenhsin, @davemorin, and @vincentkoc.
**UI:** Keep key-free providers such as Parallel Free, DuckDuckGo, Ollama, and Codex Hosted Search as explicit opt-ins instead of selecting them automatically when no API-backed provider is configured. Thanks @davemorin and @vincentkoc.
**Channel plugins:** Ship Telegram rich-message delivery and WhatsApp ACP binding support, including preserved intentional line breaks, rich prompt handoff to CLI backends, and transport fixtures for richer drafts. Thanks @obviyus, @TurboTheTurtle, @vincentkoc, @mcaxtr, and @dmorn.
...and 2 more
🐛 Bug Fixes
**Channels and delivery:** Preserve account-scoped DM channel send policy, intentional rich-message line breaks in Telegram and status output, rich Telegram final replies, rich Telegram tables and lists, Telegram thread-create CLI remapping, Feishu dynamic-agent routes after persisted binding reuse, Slack outbound `message_sent` hooks, contributed message-tool schema optionality, same-channel generated media completions, and channel chunking around surrogate pairs and Infinity limits. Thanks @yetval, @obviyus, @spacegeologist, @rishitamrakar, @liuhao1024, @lundog, @TurboTheTurtle, @yhterrance, @vincentkoc, @myrzka, @cwlong163-afk, @kumaxs, @shakkernerd, and @RewardsPal.
**CLI:** Use the selected OpenClaw OAuth/API-key auth profile in an isolated Gemini CLI runtime home, preventing ambient Google machine credentials from overriding the chosen profile. Thanks @jason-allen-oneal and @shakkernerd.
**Discord:** Give generated auto-thread titles a 60-second timeout and 4,096-token reasoning-model output budget, clamped to the selected model output cap.
Agent, cron, and Gateway runtime: mark active main sessions before restart shutdown aborts, pause yielded subagent runs whose terminal also signals abort, clamp trusted subagent thinking overrides through provider/model fallback, preserve yielded media completions, deliver channel message-tool final replies through auto-reply while hiding internal delivery hints, restore reset archive fallback reads when active async transcripts are missing, de-duplicate main-session heartbeat events, expose session identity in runtime prompts, reject unknown OpenAI agent selectors, keep generated media completions, slash-command block replies, and trajectory export commands in WebChat, and require admin privileges for HTTP session/model override surfaces. Thanks @ooiuuii, @openperf, @IWhatsskill, @masatohoshino, @CadanHu, @ZengWen-DT, @zhangguiping-xydt, @TurboTheTurtle, @oiGaDio, @aleps001, @vincentkoc, @GSL-R, @QQSHI13, @ryanhelms, @kumaxs, @steipete-oai, @hxy91819, @davemorin, and @nailujac.
**Providers and model replay:** Preserve storeless OpenAI Responses replay compatibility, recover invalid OpenAI reasoning signatures and genericized Anthropic thinking-signature replay errors, route OAuth image defaults through Codex for eligible OpenAI profiles, avoid eager tool streaming for Claude 4.5 in Copilot, quarantine unreadable and post-hook OpenAI/Anthropic-family tool schemas without broadening allowed tool choices, deliver explicit thinking-off requests to LM Studio binary-thinking models, honor profile auth for SecretRef model entries, bound model browsing, strip provider prefixes where runtimes need bare IDs, and surface nested embedding fetch failures. Thanks @snowzlm, @mmyzwl, @CarlCapital, @bek91, @Kailigithub, @vincentkoc, @rohitjavvadi, @samson910022, @nxmxbbd, @liuhao1024, @bymle, @mushuiyu886, @finchinslc, @syfvb, @lijenhsin, @crsnpalmer-art, @samson1357924, @shakkernerd, and @mlaihk.
Memory, state, diagnostics, and config: split header-too-large embedding batches, keep QMD memory search enabled in transient mode, **Fixed:** SQLite WAL on NFS volumes, preserve recovery scheduling outside stuck-session warning backoff, preserve full-reindex rollback/cache recovery, and treat raw Memory Wiki source pages as source evidence. Thanks @mushuiyu886, @TurboTheTurtle, @849261680, @gnanam1990, @TSHOGX, @vincentkoc, @arlen8411, @BrettHamlin, @zhbcher, @Takhoffman, @AFabyTWE, @davemorin, and @zhuyankarl.
**Plugins and updates:** Repair missing required platform packages during managed plugin installs and updates, including omitted Codex platform binaries.
**Dependencies:** Update Hono to 4.12.25 so published OpenClaw and ACPX packages use the patched runtime.
**Updates:** Avoid a false downgrade prompt when the latest tag cannot resolve. Thanks @Andy312432 and @vincentkoc.
**Richer channel delivery:** Telegram and WhatsApp are less brittle: Telegram renders structured text with tables, lists, expandable blockquotes, preserved intentional line breaks, and CLI-backed replies, while WhatsApp now honors configured ACP bindings. Thanks @obviyus, @vincentkoc, @jzakirov, @spacegeologist, @TurboTheTurtle, @mcaxtr, @myrzka, and @dmorn.
**More reliable agent runs:** account-scoped DM sends, generated media completions, auto-reply message-tool final replies, reset archive fallback reads, restart shutdown aborts, yielded subagent pauses, and session identity prompts all stay on the correct recovery path. Thanks @yetval, @TurboTheTurtle, @masatohoshino, @CadanHu, @vincentkoc, @ooiuuii, @openperf, @zhangguiping-xydt, @QQSHI13, @kumaxs, and @aleps001.
**Safer model routing:** new GLM-5.2 and Claude Haiku 4.5 catalog support arrives with normalized provider IDs, managed SecretRef auth, bounded model browsing, and safer OpenAI/Anthropic tool-schema recovery. Thanks @arkyu2077, @liuhao1024, @lijenhsin, @rohitjavvadi, @samson910022, @maaron34, @syfvb, and @samson1357924.
**Useful usage footers:** `/usage` and reply payload hooks now have a native full footer renderer, default template, fixed-decimal formatting, credential-aware limits, **Improved:** Partial-count handling, and warnings for broken templates instead of silent bad output.
**Predictable web search defaults:** key-free providers such as Parallel Free, DuckDuckGo, Ollama, and Codex Hosted Search remain explicit opt-ins rather than surprising automatic fallbacks. Thanks @davemorin and @vincentkoc.
**Calmer UI and mobile sessions:** workspace files start collapsed, WebChat backscroll survives streaming, the desktop session picker remains interactive, reset arguments survive dispatch, and iOS reconnects stale foreground Gateways. Thanks @shakkernerd, @TurboTheTurtle, @NianJiuZst, @zhouhe-xydt, @Solvely-Colin, @MaBeitian, @vincentkoc, @Chang2020618, and @DrtyMorty.
**Resilient memory and state:** oversized OpenAI embedding batches split before 431s, QMD search stays available in transient mode, SQLite avoids WAL on NFS volumes, and full reindexes preserve rollback/cache recovery. Thanks @mushuiyu886, @BrettHamlin, @zhbcher, @TurboTheTurtle, @Takhoffman, @849261680, @TSHOGX, @vincentkoc, and @AFabyTWE.
**Providers/models:** Add GLM-5.2 support and Claude Haiku 4.5 catalog entries while keeping provider-qualified model IDs normalized across OpenRouter and Google Vertex paths. Thanks @arkyu2077, @liuhao1024, @bymle, @maaron34, @lijenhsin, @davemorin, and @vincentkoc.
**UI:** Keep key-free providers such as Parallel Free, DuckDuckGo, Ollama, and Codex Hosted Search as explicit opt-ins instead of selecting them automatically when no API-backed provider is configured. Thanks @davemorin and @vincentkoc.
**Channel plugins:** Ship Telegram rich-message delivery and WhatsApp ACP binding support, including preserved intentional line breaks, rich prompt handoff to CLI backends, and transport fixtures for richer drafts. Thanks @obviyus, @TurboTheTurtle, @vincentkoc, @mcaxtr, and @dmorn.
...and 2 more
🐛 Bug Fixes
**Channels and delivery:** Preserve account-scoped DM channel send policy, intentional rich-message line breaks in Telegram and status output, rich Telegram final replies, rich Telegram tables and lists, Telegram thread-create CLI remapping, Feishu dynamic-agent routes after persisted binding reuse, Slack outbound `message_sent` hooks, contributed message-tool schema optionality, same-channel generated media completions, and channel chunking around surrogate pairs and Infinity limits. Thanks @yetval, @obviyus, @spacegeologist, @rishitamrakar, @liuhao1024, @lundog, @TurboTheTurtle, @yhterrance, @vincentkoc, @myrzka, @cwlong163-afk, @kumaxs, @shakkernerd, and @RewardsPal.
**CLI:** Use the selected OpenClaw OAuth/API-key auth profile in an isolated Gemini CLI runtime home, preventing ambient Google machine credentials from overriding the chosen profile. Thanks @jason-allen-oneal and @shakkernerd.
**Discord:** Give generated auto-thread titles a 60-second timeout and 4,096-token reasoning-model output budget, clamped to the selected model output cap.
Agent, cron, and Gateway runtime: mark active main sessions before restart shutdown aborts, pause yielded subagent runs whose terminal also signals abort, clamp trusted subagent thinking overrides through provider/model fallback, preserve yielded media completions, deliver channel message-tool final replies through auto-reply while hiding internal delivery hints, restore reset archive fallback reads when active async transcripts are missing, de-duplicate main-session heartbeat events, expose session identity in runtime prompts, reject unknown OpenAI agent selectors, keep generated media completions, slash-command block replies, and trajectory export commands in WebChat, and require admin privileges for HTTP session/model override surfaces. Thanks @ooiuuii, @openperf, @IWhatsskill, @masatohoshino, @CadanHu, @ZengWen-DT, @zhangguiping-xydt, @TurboTheTurtle, @oiGaDio, @aleps001, @vincentkoc, @GSL-R, @QQSHI13, @ryanhelms, @kumaxs, @steipete-oai, @hxy91819, @davemorin, and @nailujac.
**Providers and model replay:** Preserve storeless OpenAI Responses replay compatibility, recover invalid OpenAI reasoning signatures and genericized Anthropic thinking-signature replay errors, route OAuth image defaults through Codex for eligible OpenAI profiles, avoid eager tool streaming for Claude 4.5 in Copilot, quarantine unreadable and post-hook OpenAI/Anthropic-family tool schemas without broadening allowed tool choices, deliver explicit thinking-off requests to LM Studio binary-thinking models, honor profile auth for SecretRef model entries, bound model browsing, strip provider prefixes where runtimes need bare IDs, and surface nested embedding fetch failures. Thanks @snowzlm, @mmyzwl, @CarlCapital, @bek91, @Kailigithub, @vincentkoc, @rohitjavvadi, @samson910022, @nxmxbbd, @liuhao1024, @bymle, @mushuiyu886, @finchinslc, @syfvb, @lijenhsin, @crsnpalmer-art, @samson1357924, @shakkernerd, and @mlaihk.
Memory, state, diagnostics, and config: split header-too-large embedding batches, keep QMD memory search enabled in transient mode, **Fixed:** SQLite WAL on NFS volumes, preserve recovery scheduling outside stuck-session warning backoff, preserve full-reindex rollback/cache recovery, and treat raw Memory Wiki source pages as source evidence. Thanks @mushuiyu886, @TurboTheTurtle, @849261680, @gnanam1990, @TSHOGX, @vincentkoc, @arlen8411, @BrettHamlin, @zhbcher, @Takhoffman, @AFabyTWE, @davemorin, and @zhuyankarl.
**Plugins and updates:** Repair missing required platform packages during managed plugin installs and updates, including omitted Codex platform binaries.
**Dependencies:** Update Hono to 4.12.25 so published OpenClaw and ACPX packages use the patched runtime.
**Updates:** Avoid a false downgrade prompt when the latest tag cannot resolve. Thanks @Andy312432 and @vincentkoc.
**Durable channel replies:** Telegram polling and preview failures recover instead of ending a stream, Slack keeps delivered replies in the transcript, and top-level image sends retain their intended media source. Thanks @joshavant, @TurboTheTurtle, @xydigit-sj, @ichirokyoto, @TalkingHeadsJed, and @riseandshinefutures.
**More dependable provider sessions:** SecretRef-backed profiles, configured DeepSeek transports, static model fallback, Anthropic thinking replay, and Codex prompt memory all recover cleanly instead of leaving a turn unusable. Thanks @joshavant, @rubencu, @sallyom, @marcoraepple-sys, @LINSUISHENG034, @mattsfraser, and @reginaldomarcilon.
**Kimi K2.7 Code support:** the provider catalog now includes the new Kimi coding model, with replay handling that preserves its reasoning content across tool turns. Thanks @xialonglee and @RoseKongPS.
**Safer operations:** SQLite-backed cron status, disabled heartbeat retries, Linux service updates, and external-plugin diagnosis now expose clearer, actionable state to operators. Thanks @liuhao1024, @joshavant, @brokemac79, @AaronFaby, @cdeyoung67, @ooiuuii, @shakkernerd, @A1fred-AI, and @hanyizuo.
**Guarded skill installs:** Skill Workshop support-file targets now go through trusted lifecycle writes instead of unbounded filesystem updates, while ClawHub package checks stay on the current release path.
Skills and plugin workflows now permit trusted Skill Workshop support-file targets only through guarded lifecycle writes, and package publishing uses the current ClawHub plugin checks. Thanks @abnershang and @vincentkoc.
**Providers:** Add Kimi K2.7 Code support.
🐛 Bug Fixes
**Channels and delivery:** Recover Telegram preview and polling failures, retain Slack final replies in transcripts, preserve top-level outbound image parameters, and make channel-action result pages available to callers. Thanks @joshavant, @TurboTheTurtle, @xydigit-sj, @fuller-stack-dev, @TalkingHeadsJed, and @riseandshinefutures.
**Agent/provider reliability:** Preserve configured model transport/auth resolution, fail closed for unsupported CLI-backed `/btw` fallback, continue after source message-tool replies, repair Anthropic thinking replay, and keep Codex memory prompts registered. Thanks @joshavant, @rubencu, @sallyom, @marcoraepple-sys, @wangwllu, @elyalvarado, and @reginaldomarcilon.
**Operations:** Make cron and daemon status resilient, preserve disabled heartbeat one-shot retries, hand off Linux service auto-updates, and keep lifecycle timeout cleanup alive after leader exit. Thanks @liuhao1024, @joshavant, @RomneyDa, @AaronFaby, @A1fred-AI, and @hanyizuo.
**PR #92144** fix(cron): report SQLite storage path in cron.status instead of legacy jobs.json. Related #91766. Thanks @liuhao1024 and @AaronFaby.
**PR #92175** fix(channel): harden local setup trust.
**PR #91528** **Fixed:** #73837: stop after failed Node package installs. Thanks @mushuiyu886 and @ItsMeForLua.
**PR #91561** fix(wizard): report keyless web_search providers as ready, not missing a key.
**PR #92073** fix: handle explicit silent assistant replies. Related #92038. Thanks @sallyom and @vultusv.
**PR #91311** You can now Skill Workshop apply through trusted skill symlinks.
**Tighter security boundaries:** transcript, sandbox, MCP, browser, channel, and exec-approval paths now fail closed around unsafe access, timed-out approvals, and malformed boundary input. Thanks @joshavant, @pgondhi987, @mmaps, @eleqtrizit, @drobison00, @vincentkoc, and @devinkuhn.
**Reliable Telegram delivery:** account-scoped topics route to the correct agent, streamed text survives tool calls, callbacks and draft chunks stay coherent, and unauthorized DM text does not enter cache or prompt context. Thanks @codysai001, @alexzhu0, @snowzlm, @obviyus, @sallyom, @AbdelftahZowail, @producedbysavant, @shakkernerd, @vincentkoc, and @BSG2000.
**iMessage stays connected:** always-on inbound recovery, durable echo markers, block streaming, idle approval discovery, and outbound transport now survive restarts and idle periods. Thanks @omarshahine, @jmissig, @dwonshin, @colmbrogan, @vincentkoc, and @TurboTheTurtle.
****Improved:** Browser and MCP connectivity:** existing browser sessions, CDP/WebSocket discovery, default-profile URLs, OAuth/SSE transport, and tool schemas now connect through clearer, safer paths. Thanks @pgondhi987, @anagnorisis2peripeteia, @eleqtrizit, @LiuwqGit, @lifuyue, @marcusbsorensen, @cursoragent, @vincentkoc, @849261680, and @mgrandau.
****Performance:** Faster first replies:** Control UI startup no longer waits on broad model loading, while cached metadata, lazy slash-command work, and first-event tracing make slow initial responses visible. Thanks @vincentkoc and @BSG2000.
**Broader provider support:** OpenRouter OAuth and Claude Fable 5 land alongside correct Codex compaction ownership, local-model execution, normalized tool progress, and Gemma 4 reasoning replay. Thanks @Patrick-Erichsen, @joshavant, @bdjben, @Coder-Wangyankun, @vincentkoc, @bfox55, @shakkernerd, and @NOVA-Openclaw.
**CLI:** Emit Claude CLI commentary progress events and bridge inter-tool commentary into channel progress without exposing internal protocol scaffolding. Thanks @anagnorisis2peripeteia, @AbdelftahZowail, @kentuscn, and @vincentkoc.
**Observability:** Allow trusted diagnostics channels to capture tool input/output content, add first-assistant-event traces, and warn on slow initial replies. Thanks @amknight, @mjunaidca, and @vincentkoc.
**Plugins/ClawHub:** Dogfood reusable package publishing, let dry runs skip publish approval, allow declared installed trusted hooks, report managed plugin version drift, and warn instead of failing on retired Skill Workshop configuration. Thanks @Patrick-Erichsen, @brokemac79, @lonexreb, @rogerallen1, @vincentkoc, and @ryanhelms.
**Memory/providers:** Move the local llama.cpp runtime into its provider plugin, batch embeddings across files, persist the agent model catalog cache, and keep QMD JSON search one-shot while filtering stale REM recall previews. Thanks @osolmaz, @mushuiyu886, @ai-hpc, @TurboTheTurtle, @jalehman, @hartmark, @vincentkoc, @rudi193-cmd, @Peilsender, and @xpysgdhr.
...and 2 more
🐛 Bug Fixes
**Agent/session recovery:** Drop stale approval follow-ups after session rebind, remove drained reply-queue items by identity, recover stale main and visible replies, preserve Codex context-engine compaction ownership, lower the default compaction timeout to 180 seconds while respecting explicit configuration, and keep provider-failure terminal lifecycle state correct. Thanks @openperf, @yetval, @joshavant, @wangmiao0668000666, @TurboTheTurtle, @two3pro, @velvet-shark, @sallyom, @849261680, @vincentkoc, @Tony-ooo, @Jerry-Xin, @olveww-dot, and @nikhilmaddirala.
User-visible content boundaries: suppress Codex/Harmony protocol artifacts, neutralize browser and LanceDB memory media directives, redact transcript images, and preserve native `/compact` replies through source suppression. Thanks @joelnishanth, @pgondhi987, @joshavant, @snowzlm, @reslp, @vincentkoc, and @devinkuhn.
**Channel delivery:** Keep WhatsApp captured replies attached to the successor controller after restart, retry Feishu rate limits, preserve Mattermost thread replies, canonicalize LINE webhook paths, restore Discord reply hydration and runtime timeout exports, and show OpenAI Realtime WebRTC assistant transcripts. Thanks @itsuzef, @ladygege, @jacobtomlinson, @fuller-stack-dev, @shushushv, @mcaxtr, @AxelHu, @vincentkoc, @marshallm-create, @sliverp, and @dahifi.
**Cron:** Cancel active task runs cleanly, preserve terminal timeout/cancel state, and recover no-deliver tool warnings instead of silently losing the outcome.
**🔒 Gateway/config/auth:** Share the approval runtime socket token, replace arrays explicitly in `config.patch`, skip the deleted-agent guard only for valid ACP harness sessions, surface headless LaunchAgent state, verify SQLite auth migration before cleanup, and arm QMD startup maintenance. Thanks @fuller-stack-dev, @scotthuang, @joshavant, @velvet-shark, @vincentkoc, and @dahifi.
**Providers/Codex:** Clarify quota errors, restore the Codex synthetic usage line, canonicalize Codex protocol assets, require API-key auth for realtime voice, normalize ACP model refs, preserve Gemma 4 `reasoning_content`, and avoid guardian review for local models. Thanks @hxy91819, @brokemac79, @RomneyDa, @joshavant, @Coder-Wangyankun, @vincentkoc, @bfox55, @shakkernerd, and @sergiopesch.
**UI:** Recover package Gateway restarts after refresh failure, expose plugin convergence repair, fall back to Corepack in PATH-less pnpm environments, seed the correct Docker store packages, and keep ClawHub dry-run and publish paths reusable. Thanks @fuller-stack-dev, @sallyom, @Patrick-Erichsen, @vincentkoc, and @laurenceputra.
**UI:** Require explicit user intent before opening chat sessions and drain restored chat queues after session switches. Thanks @TurboTheTurtle, @Takhoffman, and @zdwalter.
**Android:** Avoid the `dataSync` foreground-service type for persistent nodes.
**Native hooks:** Bound relay lifetimes so abandoned native hook connections cannot linger indefinitely. Thanks @joshavant and @clem-git.
**Tighter security boundaries:** transcript, sandbox, MCP, browser, channel, and exec-approval paths now fail closed around unsafe access, timed-out approvals, and malformed boundary input. Thanks @joshavant, @pgondhi987, @mmaps, @eleqtrizit, @drobison00, @vincentkoc, and @devinkuhn.
**Reliable Telegram delivery:** account-scoped topics route to the correct agent, streamed text survives tool calls, callbacks and draft chunks stay coherent, and unauthorized DM text does not enter cache or prompt context. Thanks @codysai001, @alexzhu0, @snowzlm, @obviyus, @sallyom, @AbdelftahZowail, @producedbysavant, @shakkernerd, @vincentkoc, and @BSG2000.
**iMessage stays connected:** always-on inbound recovery, durable echo markers, block streaming, idle approval discovery, and outbound transport now survive restarts and idle periods. Thanks @omarshahine, @jmissig, @dwonshin, @colmbrogan, @vincentkoc, and @TurboTheTurtle.
****Improved:** Browser and MCP connectivity:** existing browser sessions, CDP/WebSocket discovery, default-profile URLs, OAuth/SSE transport, and tool schemas now connect through clearer, safer paths. Thanks @pgondhi987, @anagnorisis2peripeteia, @eleqtrizit, @LiuwqGit, @lifuyue, @marcusbsorensen, @cursoragent, @vincentkoc, @849261680, and @mgrandau.
****Performance:** Faster first replies:** Control UI startup no longer waits on broad model loading, while cached metadata, lazy slash-command work, and first-event tracing make slow initial responses visible. Thanks @vincentkoc and @BSG2000.
**Broader provider support:** OpenRouter OAuth and Claude Fable 5 land alongside correct Codex compaction ownership, local-model execution, normalized tool progress, and Gemma 4 reasoning replay. Thanks @Patrick-Erichsen, @joshavant, @bdjben, @Coder-Wangyankun, @vincentkoc, @bfox55, @shakkernerd, and @NOVA-Openclaw.
**CLI:** Emit Claude CLI commentary progress events and bridge inter-tool commentary into channel progress without exposing internal protocol scaffolding. Thanks @anagnorisis2peripeteia, @AbdelftahZowail, @kentuscn, and @vincentkoc.
**Observability:** Allow trusted diagnostics channels to capture tool input/output content, add first-assistant-event traces, and warn on slow initial replies. Thanks @amknight, @mjunaidca, and @vincentkoc.
**Plugins/ClawHub:** Dogfood reusable package publishing, let dry runs skip publish approval, allow declared installed trusted hooks, report managed plugin version drift, and warn instead of failing on retired Skill Workshop configuration. Thanks @Patrick-Erichsen, @brokemac79, @lonexreb, @rogerallen1, @vincentkoc, and @ryanhelms.
**Memory/providers:** Move the local llama.cpp runtime into its provider plugin, batch embeddings across files, persist the agent model catalog cache, and keep QMD JSON search one-shot while filtering stale REM recall previews. Thanks @osolmaz, @mushuiyu886, @ai-hpc, @TurboTheTurtle, @jalehman, @hartmark, @vincentkoc, @rudi193-cmd, @Peilsender, and @xpysgdhr.
...and 2 more
🐛 Bug Fixes
**Agent/session recovery:** Drop stale approval follow-ups after session rebind, remove drained reply-queue items by identity, recover stale main and visible replies, preserve Codex context-engine compaction ownership, lower the default compaction timeout to 180 seconds while respecting explicit configuration, and keep provider-failure terminal lifecycle state correct. Thanks @openperf, @yetval, @joshavant, @wangmiao0668000666, @TurboTheTurtle, @two3pro, @velvet-shark, @sallyom, @849261680, @vincentkoc, @Tony-ooo, @Jerry-Xin, @olveww-dot, and @nikhilmaddirala.
User-visible content boundaries: suppress Codex/Harmony protocol artifacts, neutralize browser and LanceDB memory media directives, redact transcript images, and preserve native `/compact` replies through source suppression. Thanks @joelnishanth, @pgondhi987, @joshavant, @snowzlm, @reslp, @vincentkoc, and @devinkuhn.
**Channel delivery:** Keep WhatsApp captured replies attached to the successor controller after restart, retry Feishu rate limits, preserve Mattermost thread replies, canonicalize LINE webhook paths, restore Discord reply hydration and runtime timeout exports, and show OpenAI Realtime WebRTC assistant transcripts. Thanks @itsuzef, @ladygege, @jacobtomlinson, @fuller-stack-dev, @shushushv, @mcaxtr, @AxelHu, @vincentkoc, @marshallm-create, @sliverp, and @dahifi.
**Cron:** Cancel active task runs cleanly, preserve terminal timeout/cancel state, and recover no-deliver tool warnings instead of silently losing the outcome.
**🔒 Gateway/config/auth:** Share the approval runtime socket token, replace arrays explicitly in `config.patch`, skip the deleted-agent guard only for valid ACP harness sessions, surface headless LaunchAgent state, verify SQLite auth migration before cleanup, and arm QMD startup maintenance. Thanks @fuller-stack-dev, @scotthuang, @joshavant, @velvet-shark, @vincentkoc, and @dahifi.
**Providers/Codex:** Clarify quota errors, restore the Codex synthetic usage line, canonicalize Codex protocol assets, require API-key auth for realtime voice, normalize ACP model refs, preserve Gemma 4 `reasoning_content`, and avoid guardian review for local models. Thanks @hxy91819, @brokemac79, @RomneyDa, @joshavant, @Coder-Wangyankun, @vincentkoc, @bfox55, @shakkernerd, and @sergiopesch.
**UI:** Recover package Gateway restarts after refresh failure, expose plugin convergence repair, fall back to Corepack in PATH-less pnpm environments, seed the correct Docker store packages, and keep ClawHub dry-run and publish paths reusable. Thanks @fuller-stack-dev, @sallyom, @Patrick-Erichsen, @vincentkoc, and @laurenceputra.
**UI:** Require explicit user intent before opening chat sessions and drain restored chat queues after session switches. Thanks @TurboTheTurtle, @Takhoffman, and @zdwalter.
**Android:** Avoid the `dataSync` foreground-service type for persistent nodes.
**Native hooks:** Bound relay lifetimes so abandoned native hook connections cannot linger indefinitely. Thanks @joshavant and @clem-git.
**Tighter security boundaries:** transcript, sandbox, MCP, browser, channel, and exec-approval paths now fail closed around unsafe access, timed-out approvals, and malformed boundary input. Thanks @joshavant, @pgondhi987, @mmaps, @eleqtrizit, @drobison00, @vincentkoc, and @devinkuhn.
**Reliable Telegram delivery:** account-scoped topics route to the correct agent, streamed text survives tool calls, callbacks and draft chunks stay coherent, and unauthorized DM text does not enter cache or prompt context. Thanks @codysai001, @alexzhu0, @snowzlm, @obviyus, @sallyom, @AbdelftahZowail, @producedbysavant, @shakkernerd, @vincentkoc, and @BSG2000.
**iMessage stays connected:** always-on inbound recovery, durable echo markers, block streaming, idle approval discovery, and outbound transport now survive restarts and idle periods. Thanks @omarshahine, @jmissig, @dwonshin, @colmbrogan, @vincentkoc, and @TurboTheTurtle.
****Improved:** Browser and MCP connectivity:** existing browser sessions, CDP/WebSocket discovery, default-profile URLs, OAuth/SSE transport, and tool schemas now connect through clearer, safer paths. Thanks @pgondhi987, @anagnorisis2peripeteia, @eleqtrizit, @LiuwqGit, @lifuyue, @marcusbsorensen, @cursoragent, @vincentkoc, @849261680, and @mgrandau.
****Performance:** Faster first replies:** Control UI startup no longer waits on broad model loading, while cached metadata, lazy slash-command work, and first-event tracing make slow initial responses visible. Thanks @vincentkoc and @BSG2000.
**Broader provider support:** OpenRouter OAuth and Claude Fable 5 land alongside correct Codex compaction ownership, local-model execution, normalized tool progress, and Gemma 4 reasoning replay. Thanks @Patrick-Erichsen, @joshavant, @bdjben, @Coder-Wangyankun, @vincentkoc, @bfox55, @shakkernerd, and @NOVA-Openclaw.
**CLI:** Emit Claude CLI commentary progress events and bridge inter-tool commentary into channel progress without exposing internal protocol scaffolding. Thanks @anagnorisis2peripeteia, @AbdelftahZowail, @kentuscn, and @vincentkoc.
**Observability:** Allow trusted diagnostics channels to capture tool input/output content, add first-assistant-event traces, and warn on slow initial replies. Thanks @amknight, @mjunaidca, and @vincentkoc.
**Plugins/ClawHub:** Dogfood reusable package publishing, let dry runs skip publish approval, allow declared installed trusted hooks, report managed plugin version drift, and warn instead of failing on retired Skill Workshop configuration. Thanks @Patrick-Erichsen, @brokemac79, @lonexreb, @rogerallen1, @vincentkoc, and @ryanhelms.
**Memory/providers:** Move the local llama.cpp runtime into its provider plugin, batch embeddings across files, persist the agent model catalog cache, and keep QMD JSON search one-shot while filtering stale REM recall previews. Thanks @osolmaz, @mushuiyu886, @ai-hpc, @TurboTheTurtle, @jalehman, @hartmark, @vincentkoc, @rudi193-cmd, @Peilsender, and @xpysgdhr.
...and 2 more
🐛 Bug Fixes
**Agent/session recovery:** Drop stale approval follow-ups after session rebind, remove drained reply-queue items by identity, recover stale main and visible replies, preserve Codex context-engine compaction ownership, lower the default compaction timeout to 180 seconds while respecting explicit configuration, and keep provider-failure terminal lifecycle state correct. Thanks @openperf, @yetval, @joshavant, @wangmiao0668000666, @TurboTheTurtle, @two3pro, @velvet-shark, @sallyom, @849261680, @vincentkoc, @Tony-ooo, @Jerry-Xin, @olveww-dot, and @nikhilmaddirala.
User-visible content boundaries: suppress Codex/Harmony protocol artifacts, neutralize browser and LanceDB memory media directives, redact transcript images, and preserve native `/compact` replies through source suppression. Thanks @joelnishanth, @pgondhi987, @joshavant, @snowzlm, @reslp, @vincentkoc, and @devinkuhn.
**Channel delivery:** Keep WhatsApp captured replies attached to the successor controller after restart, retry Feishu rate limits, preserve Mattermost thread replies, canonicalize LINE webhook paths, restore Discord reply hydration and runtime timeout exports, and show OpenAI Realtime WebRTC assistant transcripts. Thanks @itsuzef, @ladygege, @jacobtomlinson, @fuller-stack-dev, @shushushv, @mcaxtr, @AxelHu, @vincentkoc, @marshallm-create, @sliverp, and @dahifi.
**Cron:** Cancel active task runs cleanly, preserve terminal timeout/cancel state, and recover no-deliver tool warnings instead of silently losing the outcome.
**🔒 Gateway/config/auth:** Share the approval runtime socket token, replace arrays explicitly in `config.patch`, skip the deleted-agent guard only for valid ACP harness sessions, surface headless LaunchAgent state, verify SQLite auth migration before cleanup, and arm QMD startup maintenance. Thanks @fuller-stack-dev, @scotthuang, @joshavant, @velvet-shark, @vincentkoc, and @dahifi.
**Providers/Codex:** Clarify quota errors, restore the Codex synthetic usage line, canonicalize Codex protocol assets, require API-key auth for realtime voice, normalize ACP model refs, preserve Gemma 4 `reasoning_content`, and avoid guardian review for local models. Thanks @hxy91819, @brokemac79, @RomneyDa, @joshavant, @Coder-Wangyankun, @vincentkoc, @bfox55, @shakkernerd, and @sergiopesch.
**UI:** Recover package Gateway restarts after refresh failure, expose plugin convergence repair, fall back to Corepack in PATH-less pnpm environments, seed the correct Docker store packages, and keep ClawHub dry-run and publish paths reusable. Thanks @fuller-stack-dev, @sallyom, @Patrick-Erichsen, @vincentkoc, and @laurenceputra.
**UI:** Require explicit user intent before opening chat sessions and drain restored chat queues after session switches. Thanks @TurboTheTurtle, @Takhoffman, and @zdwalter.
**Android:** Avoid the `dataSync` foreground-service type for persistent nodes.
**Native hooks:** Bound relay lifetimes so abandoned native hook connections cannot linger indefinitely. Thanks @joshavant and @clem-git.
**Safer channel output:** QQBot strips model reasoning and thinking scaffolding before native delivery, so users see the final answer rather than raw internal markup. Thanks @openperf, @dygg2001, @Takhoffman, and @vincentkoc.
**MCP results no longer poison sessions:** `resource_link`, audio, malformed images, and future non-text blocks are normalized before provider conversion, avoiding Anthropic 400s and broken follow-up history. Thanks @RanSHammer, @849261680, and @Takhoffman.
**Anthropic extended thinking recovers after restarts:** prompt-cache expiry and early-signature failures wait for a real message start, so they reach the existing retry path rather than ending a turn. Thanks @openperf, @MIHHHMIH, @Takhoffman, @vincentkoc, @itsuzef, and @mcaxtr.
**Parallel web search is bundled:** API-key discovery, guarded endpoint handling, cache-safe session IDs, onboarding, and documentation make it a first-class search provider.
****Improved:** Matrix and Vertex reliability:** Matrix voice notes and threaded conversations retain their context, while Google Vertex ADC models regain static catalog and runtime resolution. Thanks @849261680, @frankdierolf, and @paulogogs.
**Safer upgrades and restarts:** cron JSON stores migrate before runtime, service env placeholders **Fixed:** Masking secrets, and macOS avoids unnecessary direct-Gateway reconnect churn. Thanks @MonkeyLeeT, @sallyom, @vrurg, @wlassalle724, @jalehman, @Takhoffman, @zhangguiping-xydt, @joshavant, and @vincentkoc.
**Search/providers:** Add the Parallel bundled web-search plugin, registration contracts, onboarding integration, and guarded `api.parallel.ai/v1/search` support. Thanks @NormallyGaussian and @vincentkoc.
**Matrix/channels:** Add voice-message preflight and thread-aware read/reply behavior.
**Skills/ClawHub:** Install ClawHub skills backed by GitHub repositories through the resolved install API, download the pinned GitHub commit, keep install-policy checks, and report install telemetry after success. Thanks @Patrick-Erichsen, @vincentkoc, @itsuzef, and @mcaxtr.
**Google Chat/channels:** Add native approval card actions and click handling so Google Chat approvals use platform-native cards instead of generic message flow.
...and 3 more
🐛 Bug Fixes
**Agents:** `sessions_send` now honors an explicit `sessionKey` when stale label metadata is also present, and denied session-id sends no longer echo the resolved canonical session key. Fixes #64699; refs #74009 and #41199 Thanks @Mintalix, @RevisitMoon, @Mocha-s, @chouxiaozi1989, @sunxq1017-hash, @vincentkoc, and @joshavant.
**Channel content boundaries:** QQBot now strips reasoning/thinking tags before sending, preserving final answers while hiding internal model narration from users. Thanks @openperf, @Takhoffman, @dygg2001, and @vincentkoc.
**Agents/MCP/providers:** Coerce non-text/image MCP tool-result blocks before they reach provider converters, preserving valid images and turning richer MCP content into text instead of malformed image blocks. Thanks @RanSHammer, @849261680, @Takhoffman, @vincentkoc, and @LiuwqGit.
**Anthropic/Codex/ACP/agent recovery:** Defer Anthropic stream start events until `message_start`, strip stale compaction thinking signatures before Anthropic replay, detect unsigned thinking-only stalls, refresh prompt fences after compaction writes, reject empty completion handoffs, preserve parent streaming-off overrides/shared progress commentary, forward heartbeat metadata to context-engine hooks, and cover Codex session/thread migration edge cases. Thanks @openperf, @100yenadmin, @ooiuuii, @johnib, @Takhoffman, @MIHHHMIH, @dexiosmb, @zenglingbiao, @jalehman, @huangxun375-stack, @holgergruenhagen, @vincentkoc, @joshavant, and @ArthurusDent.
**Provider/model resolution:** Preserve Google Vertex ADC auth markers in generated catalogs, re-probe a single-provider primary after cooldown, share Codex model visibility, fail closed for unknown model auth, preserve Codex alias availability, keep unresolved profile refs unknown, and avoid resolving auth while listing models. Thanks @849261680, @paulogogs, @brtkwr, and @vincentkoc.
**Gateway/macOS/mobile:** Avoid duplicate Gateway probe warnings by identity, rate-limit node pairing requests while preserving paired-node reconnects, keep macOS node mode on a healthy direct Gateway session, keep iOS diagnostics and gateway rows reachable, and avoid Linux ARM Gradle resource tasks during Android builds. Thanks @giodl73-repo, @vrurg, @pgondhi987, @Takhoffman, @zhangguiping-xydt, @vincentkoc, @joshavant, and @shakkernerd.
TUI/chat/Workboard/auto-reply: optimistic user messages stay stable across stale history reloads, runId reassignment, and abort windows instead of disappearing, jumping, or lingering as ghost rows; Workboard stale lifecycle bulk updates no longer overwrite newer status/provenance; message-tool sends now count as delivery. Thanks @RomneyDa, @BunsDev, @kzclaw, @mcaxtr, @vincentkoc, @joshavant, and @LiuwqGit.
**Cron/update/service env:** Doctor config preflight now migrates legacy cron JSON stores into SQLite before runtime reads, service env planning skips unresolved placeholders that would mask state-dir `.env` values, and session transcript rewrites keep registry markers/discriminants consistent. Thanks @MonkeyLeeT, @sallyom, @Kvikkulf, @jalehman, @wlassalle724, @shakkernerd, and @vincentkoc.
**🔒 Security/config/tooling:** Guard MCP HTTP redirects, protect global agent config defaults, and keep malformed operational limits bounded and explicit. Thanks @pgondhi987, @vincentkoc, and @joshavant.
**Channels:** WhatsApp restarts when per-account config changes, bounds background startup waits, closes failed sockets, and preserves reconnect behavior; Mattermost slash commands keep their state on `globalThis`; Feishu streaming cards preserve full merged content; voice-call tracks Twilio streams after connect; ClickClack reply tools respect `toolsAllow`. Thanks @MukundaKatta, @mcaxtr, @infoanton, @mushuiyu886, @sahibzada-allahyar, @borntobefree2-cmyk, @Takhoffman, @mmaps, @MMMMSSSS8899, @ly85206559, @sliverp, @vincentkoc, @joshavant, @LiuwqGit, @itsuzef, @wjm7220, and @donkeykong91.
**Safer channel output:** QQBot strips model reasoning and thinking scaffolding before native delivery, so users see the final answer rather than raw internal markup. Thanks @openperf, @dygg2001, @Takhoffman, and @vincentkoc.
**MCP results no longer poison sessions:** `resource_link`, audio, malformed images, and future non-text blocks are normalized before provider conversion, avoiding Anthropic 400s and broken follow-up history. Thanks @RanSHammer, @849261680, and @Takhoffman.
**Anthropic extended thinking recovers after restarts:** prompt-cache expiry and early-signature failures wait for a real message start, so they reach the existing retry path rather than ending a turn. Thanks @openperf, @MIHHHMIH, @Takhoffman, @vincentkoc, @itsuzef, and @mcaxtr.
**Parallel web search is bundled:** API-key discovery, guarded endpoint handling, cache-safe session IDs, onboarding, and documentation make it a first-class search provider.
****Improved:** Matrix and Vertex reliability:** Matrix voice notes and threaded conversations retain their context, while Google Vertex ADC models regain static catalog and runtime resolution. Thanks @849261680, @frankdierolf, and @paulogogs.
**Safer upgrades and restarts:** cron JSON stores migrate before runtime, service env placeholders **Fixed:** Masking secrets, and macOS avoids unnecessary direct-Gateway reconnect churn. Thanks @MonkeyLeeT, @sallyom, @vrurg, @wlassalle724, @jalehman, @Takhoffman, @zhangguiping-xydt, @joshavant, and @vincentkoc.
**Search/providers:** Add the Parallel bundled web-search plugin, registration contracts, onboarding integration, and guarded `api.parallel.ai/v1/search` support. Thanks @NormallyGaussian and @vincentkoc.
**Matrix/channels:** Add voice-message preflight and thread-aware read/reply behavior.
**Skills/ClawHub:** Install ClawHub skills backed by GitHub repositories through the resolved install API, download the pinned GitHub commit, keep install-policy checks, and report install telemetry after success. Thanks @Patrick-Erichsen, @vincentkoc, @itsuzef, and @mcaxtr.
**Google Chat/channels:** Add native approval card actions and click handling so Google Chat approvals use platform-native cards instead of generic message flow.
...and 3 more
🐛 Bug Fixes
**Agents:** `sessions_send` now honors an explicit `sessionKey` when stale label metadata is also present, and denied session-id sends no longer echo the resolved canonical session key. Fixes #64699; refs #74009 and #41199 Thanks @Mintalix, @RevisitMoon, @Mocha-s, @chouxiaozi1989, @sunxq1017-hash, @vincentkoc, and @joshavant.
**Channel content boundaries:** QQBot now strips reasoning/thinking tags before sending, preserving final answers while hiding internal model narration from users. Thanks @openperf, @Takhoffman, @dygg2001, and @vincentkoc.
**Agents/MCP/providers:** Coerce non-text/image MCP tool-result blocks before they reach provider converters, preserving valid images and turning richer MCP content into text instead of malformed image blocks. Thanks @RanSHammer, @849261680, @Takhoffman, @vincentkoc, and @LiuwqGit.
**Anthropic/Codex/ACP/agent recovery:** Defer Anthropic stream start events until `message_start`, strip stale compaction thinking signatures before Anthropic replay, detect unsigned thinking-only stalls, refresh prompt fences after compaction writes, reject empty completion handoffs, preserve parent streaming-off overrides/shared progress commentary, forward heartbeat metadata to context-engine hooks, and cover Codex session/thread migration edge cases. Thanks @openperf, @100yenadmin, @ooiuuii, @johnib, @Takhoffman, @MIHHHMIH, @dexiosmb, @zenglingbiao, @jalehman, @huangxun375-stack, @holgergruenhagen, @vincentkoc, @joshavant, and @ArthurusDent.
**Provider/model resolution:** Preserve Google Vertex ADC auth markers in generated catalogs, re-probe a single-provider primary after cooldown, share Codex model visibility, fail closed for unknown model auth, preserve Codex alias availability, keep unresolved profile refs unknown, and avoid resolving auth while listing models. Thanks @849261680, @paulogogs, @brtkwr, and @vincentkoc.
**Gateway/macOS/mobile:** Avoid duplicate Gateway probe warnings by identity, rate-limit node pairing requests while preserving paired-node reconnects, keep macOS node mode on a healthy direct Gateway session, keep iOS diagnostics and gateway rows reachable, and avoid Linux ARM Gradle resource tasks during Android builds. Thanks @giodl73-repo, @vrurg, @pgondhi987, @Takhoffman, @zhangguiping-xydt, @vincentkoc, @joshavant, and @shakkernerd.
TUI/chat/Workboard/auto-reply: optimistic user messages stay stable across stale history reloads, runId reassignment, and abort windows instead of disappearing, jumping, or lingering as ghost rows; Workboard stale lifecycle bulk updates no longer overwrite newer status/provenance; message-tool sends now count as delivery. Thanks @RomneyDa, @BunsDev, @kzclaw, @mcaxtr, @vincentkoc, @joshavant, and @LiuwqGit.
**Cron/update/service env:** Doctor config preflight now migrates legacy cron JSON stores into SQLite before runtime reads, service env planning skips unresolved placeholders that would mask state-dir `.env` values, and session transcript rewrites keep registry markers/discriminants consistent. Thanks @MonkeyLeeT, @sallyom, @Kvikkulf, @jalehman, @wlassalle724, @shakkernerd, and @vincentkoc.
**🔒 Security/config/tooling:** Guard MCP HTTP redirects, protect global agent config defaults, and keep malformed operational limits bounded and explicit. Thanks @pgondhi987, @vincentkoc, and @joshavant.
**Channels:** WhatsApp restarts when per-account config changes, bounds background startup waits, closes failed sockets, and preserves reconnect behavior; Mattermost slash commands keep their state on `globalThis`; Feishu streaming cards preserve full merged content; voice-call tracks Twilio streams after connect; ClickClack reply tools respect `toolsAllow`. Thanks @MukundaKatta, @mcaxtr, @infoanton, @mushuiyu886, @sahibzada-allahyar, @borntobefree2-cmyk, @Takhoffman, @mmaps, @MMMMSSSS8899, @ly85206559, @sliverp, @vincentkoc, @joshavant, @LiuwqGit, @itsuzef, @wjm7220, and @donkeykong91.
**Safer channel output:** QQBot strips model reasoning and thinking scaffolding before native delivery, so users see the final answer rather than raw internal markup. Thanks @openperf, @dygg2001, @Takhoffman, and @vincentkoc.
**MCP results no longer poison sessions:** `resource_link`, audio, malformed images, and future non-text blocks are normalized before provider conversion, avoiding Anthropic 400s and broken follow-up history. Thanks @RanSHammer, @849261680, and @Takhoffman.
**Anthropic extended thinking recovers after restarts:** prompt-cache expiry and early-signature failures wait for a real message start, so they reach the existing retry path rather than ending a turn. Thanks @openperf, @MIHHHMIH, @Takhoffman, @vincentkoc, @itsuzef, and @mcaxtr.
**Parallel web search is bundled:** API-key discovery, guarded endpoint handling, cache-safe session IDs, onboarding, and documentation make it a first-class search provider.
****Improved:** Matrix and Vertex reliability:** Matrix voice notes and threaded conversations retain their context, while Google Vertex ADC models regain static catalog and runtime resolution. Thanks @849261680, @frankdierolf, and @paulogogs.
**Safer upgrades and restarts:** cron JSON stores migrate before runtime, service env placeholders **Fixed:** Masking secrets, and macOS avoids unnecessary direct-Gateway reconnect churn. Thanks @MonkeyLeeT, @sallyom, @vrurg, @wlassalle724, @jalehman, @Takhoffman, @zhangguiping-xydt, @joshavant, and @vincentkoc.
**Search/providers:** Add the Parallel bundled web-search plugin, registration contracts, onboarding integration, and guarded `api.parallel.ai/v1/search` support. Thanks @NormallyGaussian and @vincentkoc.
**Matrix/channels:** Add voice-message preflight and thread-aware read/reply behavior.
**Skills/ClawHub:** Install ClawHub skills backed by GitHub repositories through the resolved install API, download the pinned GitHub commit, keep install-policy checks, and report install telemetry after success. Thanks @Patrick-Erichsen, @vincentkoc, @itsuzef, and @mcaxtr.
**Google Chat/channels:** Add native approval card actions and click handling so Google Chat approvals use platform-native cards instead of generic message flow.
...and 3 more
🐛 Bug Fixes
**Agents:** `sessions_send` now honors an explicit `sessionKey` when stale label metadata is also present, and denied session-id sends no longer echo the resolved canonical session key. Fixes #64699; refs #74009 and #41199 Thanks @Mintalix, @RevisitMoon, @Mocha-s, @chouxiaozi1989, @sunxq1017-hash, @vincentkoc, and @joshavant.
**Channel content boundaries:** QQBot now strips reasoning/thinking tags before sending, preserving final answers while hiding internal model narration from users. Thanks @openperf, @Takhoffman, @dygg2001, and @vincentkoc.
**Agents/MCP/providers:** Coerce non-text/image MCP tool-result blocks before they reach provider converters, preserving valid images and turning richer MCP content into text instead of malformed image blocks. Thanks @RanSHammer, @849261680, @Takhoffman, @vincentkoc, and @LiuwqGit.
**Anthropic/Codex/ACP/agent recovery:** Defer Anthropic stream start events until `message_start`, strip stale compaction thinking signatures before Anthropic replay, detect unsigned thinking-only stalls, refresh prompt fences after compaction writes, reject empty completion handoffs, preserve parent streaming-off overrides/shared progress commentary, forward heartbeat metadata to context-engine hooks, and cover Codex session/thread migration edge cases. Thanks @openperf, @100yenadmin, @ooiuuii, @johnib, @Takhoffman, @MIHHHMIH, @dexiosmb, @zenglingbiao, @jalehman, @huangxun375-stack, @holgergruenhagen, @vincentkoc, @joshavant, and @ArthurusDent.
**Provider/model resolution:** Preserve Google Vertex ADC auth markers in generated catalogs, re-probe a single-provider primary after cooldown, share Codex model visibility, fail closed for unknown model auth, preserve Codex alias availability, keep unresolved profile refs unknown, and avoid resolving auth while listing models. Thanks @849261680, @paulogogs, @brtkwr, and @vincentkoc.
**Gateway/macOS/mobile:** Avoid duplicate Gateway probe warnings by identity, rate-limit node pairing requests while preserving paired-node reconnects, keep macOS node mode on a healthy direct Gateway session, keep iOS diagnostics and gateway rows reachable, and avoid Linux ARM Gradle resource tasks during Android builds. Thanks @giodl73-repo, @vrurg, @pgondhi987, @Takhoffman, @zhangguiping-xydt, @vincentkoc, @joshavant, and @shakkernerd.
TUI/chat/Workboard/auto-reply: optimistic user messages stay stable across stale history reloads, runId reassignment, and abort windows instead of disappearing, jumping, or lingering as ghost rows; Workboard stale lifecycle bulk updates no longer overwrite newer status/provenance; message-tool sends now count as delivery. Thanks @RomneyDa, @BunsDev, @kzclaw, @mcaxtr, @vincentkoc, @joshavant, and @LiuwqGit.
**Cron/update/service env:** Doctor config preflight now migrates legacy cron JSON stores into SQLite before runtime reads, service env planning skips unresolved placeholders that would mask state-dir `.env` values, and session transcript rewrites keep registry markers/discriminants consistent. Thanks @MonkeyLeeT, @sallyom, @Kvikkulf, @jalehman, @wlassalle724, @shakkernerd, and @vincentkoc.
**🔒 Security/config/tooling:** Guard MCP HTTP redirects, protect global agent config defaults, and keep malformed operational limits bounded and explicit. Thanks @pgondhi987, @vincentkoc, and @joshavant.
**Channels:** WhatsApp restarts when per-account config changes, bounds background startup waits, closes failed sockets, and preserves reconnect behavior; Mattermost slash commands keep their state on `globalThis`; Feishu streaming cards preserve full merged content; voice-call tracks Twilio streams after connect; ClickClack reply tools respect `toolsAllow`. Thanks @MukundaKatta, @mcaxtr, @infoanton, @mushuiyu886, @sahibzada-allahyar, @borntobefree2-cmyk, @Takhoffman, @mmaps, @MMMMSSSS8899, @ly85206559, @sliverp, @vincentkoc, @joshavant, @LiuwqGit, @itsuzef, @wjm7220, and @donkeykong91.
**Safer channel output:** QQBot strips model reasoning and thinking scaffolding before native delivery, so users see the final answer rather than raw internal markup. Thanks @openperf, @dygg2001, @Takhoffman, and @vincentkoc.
**MCP results no longer poison sessions:** `resource_link`, audio, malformed images, and future non-text blocks are normalized before provider conversion, avoiding Anthropic 400s and broken follow-up history. Thanks @RanSHammer, @849261680, and @Takhoffman.
**Anthropic extended thinking recovers after restarts:** prompt-cache expiry and early-signature failures wait for a real message start, so they reach the existing retry path rather than ending a turn. Thanks @openperf, @MIHHHMIH, @Takhoffman, @vincentkoc, @itsuzef, and @mcaxtr.
**Parallel web search is bundled:** API-key discovery, guarded endpoint handling, cache-safe session IDs, onboarding, and documentation make it a first-class search provider.
****Improved:** Matrix and Vertex reliability:** Matrix voice notes and threaded conversations retain their context, while Google Vertex ADC models regain static catalog and runtime resolution. Thanks @849261680, @frankdierolf, and @paulogogs.
**Safer upgrades and restarts:** cron JSON stores migrate before runtime, service env placeholders **Fixed:** Masking secrets, and macOS avoids unnecessary direct-Gateway reconnect churn. Thanks @MonkeyLeeT, @sallyom, @vrurg, @wlassalle724, @jalehman, @Takhoffman, @zhangguiping-xydt, @joshavant, and @vincentkoc.
**Search/providers:** Add the Parallel bundled web-search plugin, registration contracts, onboarding integration, and guarded `api.parallel.ai/v1/search` support. Thanks @NormallyGaussian and @vincentkoc.
**Matrix/channels:** Add voice-message preflight and thread-aware read/reply behavior.
**Skills/ClawHub:** Install ClawHub skills backed by GitHub repositories through the resolved install API, download the pinned GitHub commit, keep install-policy checks, and report install telemetry after success. Thanks @Patrick-Erichsen, @vincentkoc, @itsuzef, and @mcaxtr.
**Google Chat/channels:** Add native approval card actions and click handling so Google Chat approvals use platform-native cards instead of generic message flow.
...and 3 more
🐛 Bug Fixes
**Agents:** `sessions_send` now honors an explicit `sessionKey` when stale label metadata is also present, and denied session-id sends no longer echo the resolved canonical session key. Fixes #64699; refs #74009 and #41199 Thanks @Mintalix, @RevisitMoon, @Mocha-s, @chouxiaozi1989, @sunxq1017-hash, @vincentkoc, and @joshavant.
**Channel content boundaries:** QQBot now strips reasoning/thinking tags before sending, preserving final answers while hiding internal model narration from users. Thanks @openperf, @Takhoffman, @dygg2001, and @vincentkoc.
**Agents/MCP/providers:** Coerce non-text/image MCP tool-result blocks before they reach provider converters, preserving valid images and turning richer MCP content into text instead of malformed image blocks. Thanks @RanSHammer, @849261680, @Takhoffman, @vincentkoc, and @LiuwqGit.
**Anthropic/Codex/ACP/agent recovery:** Defer Anthropic stream start events until `message_start`, strip stale compaction thinking signatures before Anthropic replay, detect unsigned thinking-only stalls, refresh prompt fences after compaction writes, reject empty completion handoffs, preserve parent streaming-off overrides/shared progress commentary, forward heartbeat metadata to context-engine hooks, and cover Codex session/thread migration edge cases. Thanks @openperf, @100yenadmin, @ooiuuii, @johnib, @Takhoffman, @MIHHHMIH, @dexiosmb, @zenglingbiao, @jalehman, @huangxun375-stack, @holgergruenhagen, @vincentkoc, @joshavant, and @ArthurusDent.
**Provider/model resolution:** Preserve Google Vertex ADC auth markers in generated catalogs, re-probe a single-provider primary after cooldown, share Codex model visibility, fail closed for unknown model auth, preserve Codex alias availability, keep unresolved profile refs unknown, and avoid resolving auth while listing models. Thanks @849261680, @paulogogs, @brtkwr, and @vincentkoc.
**Gateway/macOS/mobile:** Avoid duplicate Gateway probe warnings by identity, rate-limit node pairing requests while preserving paired-node reconnects, keep macOS node mode on a healthy direct Gateway session, keep iOS diagnostics and gateway rows reachable, and avoid Linux ARM Gradle resource tasks during Android builds. Thanks @giodl73-repo, @vrurg, @pgondhi987, @Takhoffman, @zhangguiping-xydt, @vincentkoc, @joshavant, and @shakkernerd.
TUI/chat/Workboard/auto-reply: optimistic user messages stay stable across stale history reloads, runId reassignment, and abort windows instead of disappearing, jumping, or lingering as ghost rows; Workboard stale lifecycle bulk updates no longer overwrite newer status/provenance; message-tool sends now count as delivery. Thanks @RomneyDa, @BunsDev, @kzclaw, @mcaxtr, @vincentkoc, @joshavant, and @LiuwqGit.
**Cron/update/service env:** Doctor config preflight now migrates legacy cron JSON stores into SQLite before runtime reads, service env planning skips unresolved placeholders that would mask state-dir `.env` values, and session transcript rewrites keep registry markers/discriminants consistent. Thanks @MonkeyLeeT, @sallyom, @Kvikkulf, @jalehman, @wlassalle724, @shakkernerd, and @vincentkoc.
**🔒 Security/config/tooling:** Guard MCP HTTP redirects, protect global agent config defaults, and keep malformed operational limits bounded and explicit. Thanks @pgondhi987, @vincentkoc, and @joshavant.
**Channels:** WhatsApp restarts when per-account config changes, bounds background startup waits, closes failed sockets, and preserves reconnect behavior; Mattermost slash commands keep their state on `globalThis`; Feishu streaming cards preserve full merged content; voice-call tracks Twilio streams after connect; ClickClack reply tools respect `toolsAllow`. Thanks @MukundaKatta, @mcaxtr, @infoanton, @mushuiyu886, @sahibzada-allahyar, @borntobefree2-cmyk, @Takhoffman, @mmaps, @MMMMSSSS8899, @ly85206559, @sliverp, @vincentkoc, @joshavant, @LiuwqGit, @itsuzef, @wjm7220, and @donkeykong91.
**Safer channel output:** QQBot strips model reasoning and thinking scaffolding before native delivery, so users see the final answer rather than raw internal markup. Thanks @openperf, @dygg2001, @Takhoffman, and @vincentkoc.
**MCP results no longer poison sessions:** `resource_link`, audio, malformed images, and future non-text blocks are normalized before provider conversion, avoiding Anthropic 400s and broken follow-up history. Thanks @RanSHammer, @849261680, and @Takhoffman.
**Anthropic extended thinking recovers after restarts:** prompt-cache expiry and early-signature failures wait for a real message start, so they reach the existing retry path rather than ending a turn. Thanks @openperf, @MIHHHMIH, @Takhoffman, @vincentkoc, @itsuzef, and @mcaxtr.
**Parallel web search is bundled:** API-key discovery, guarded endpoint handling, cache-safe session IDs, onboarding, and documentation make it a first-class search provider.
****Improved:** Matrix and Vertex reliability:** Matrix voice notes and threaded conversations retain their context, while Google Vertex ADC models regain static catalog and runtime resolution. Thanks @849261680, @frankdierolf, and @paulogogs.
**Safer upgrades and restarts:** cron JSON stores migrate before runtime, service env placeholders **Fixed:** Masking secrets, and macOS avoids unnecessary direct-Gateway reconnect churn. Thanks @MonkeyLeeT, @sallyom, @vrurg, @wlassalle724, @jalehman, @Takhoffman, @zhangguiping-xydt, @joshavant, and @vincentkoc.
**Search/providers:** Add the Parallel bundled web-search plugin, registration contracts, onboarding integration, and guarded `api.parallel.ai/v1/search` support. Thanks @NormallyGaussian and @vincentkoc.
**Matrix/channels:** Add voice-message preflight and thread-aware read/reply behavior.
**Skills/ClawHub:** Install ClawHub skills backed by GitHub repositories through the resolved install API, download the pinned GitHub commit, keep install-policy checks, and report install telemetry after success. Thanks @Patrick-Erichsen, @vincentkoc, @itsuzef, and @mcaxtr.
**Google Chat/channels:** Add native approval card actions and click handling so Google Chat approvals use platform-native cards instead of generic message flow.
...and 3 more
🐛 Bug Fixes
**Agents:** `sessions_send` now honors an explicit `sessionKey` when stale label metadata is also present, and denied session-id sends no longer echo the resolved canonical session key. Fixes #64699; refs #74009 and #41199 Thanks @Mintalix, @RevisitMoon, @Mocha-s, @chouxiaozi1989, @sunxq1017-hash, @vincentkoc, and @joshavant.
**Channel content boundaries:** QQBot now strips reasoning/thinking tags before sending, preserving final answers while hiding internal model narration from users. Thanks @openperf, @Takhoffman, @dygg2001, and @vincentkoc.
**Agents/MCP/providers:** Coerce non-text/image MCP tool-result blocks before they reach provider converters, preserving valid images and turning richer MCP content into text instead of malformed image blocks. Thanks @RanSHammer, @849261680, @Takhoffman, @vincentkoc, and @LiuwqGit.
**Anthropic/Codex/ACP/agent recovery:** Defer Anthropic stream start events until `message_start`, strip stale compaction thinking signatures before Anthropic replay, detect unsigned thinking-only stalls, refresh prompt fences after compaction writes, reject empty completion handoffs, preserve parent streaming-off overrides/shared progress commentary, forward heartbeat metadata to context-engine hooks, and cover Codex session/thread migration edge cases. Thanks @openperf, @100yenadmin, @ooiuuii, @johnib, @Takhoffman, @MIHHHMIH, @dexiosmb, @zenglingbiao, @jalehman, @huangxun375-stack, @holgergruenhagen, @vincentkoc, @joshavant, and @ArthurusDent.
**Provider/model resolution:** Preserve Google Vertex ADC auth markers in generated catalogs, re-probe a single-provider primary after cooldown, share Codex model visibility, fail closed for unknown model auth, preserve Codex alias availability, keep unresolved profile refs unknown, and avoid resolving auth while listing models. Thanks @849261680, @paulogogs, @brtkwr, and @vincentkoc.
**Gateway/macOS/mobile:** Avoid duplicate Gateway probe warnings by identity, rate-limit node pairing requests while preserving paired-node reconnects, keep macOS node mode on a healthy direct Gateway session, keep iOS diagnostics and gateway rows reachable, and avoid Linux ARM Gradle resource tasks during Android builds. Thanks @giodl73-repo, @vrurg, @pgondhi987, @Takhoffman, @zhangguiping-xydt, @vincentkoc, @joshavant, and @shakkernerd.
TUI/chat/Workboard/auto-reply: optimistic user messages stay stable across stale history reloads, runId reassignment, and abort windows instead of disappearing, jumping, or lingering as ghost rows; Workboard stale lifecycle bulk updates no longer overwrite newer status/provenance; message-tool sends now count as delivery. Thanks @RomneyDa, @BunsDev, @kzclaw, @mcaxtr, @vincentkoc, @joshavant, and @LiuwqGit.
**Cron/update/service env:** Doctor config preflight now migrates legacy cron JSON stores into SQLite before runtime reads, service env planning skips unresolved placeholders that would mask state-dir `.env` values, and session transcript rewrites keep registry markers/discriminants consistent. Thanks @MonkeyLeeT, @sallyom, @Kvikkulf, @jalehman, @wlassalle724, @shakkernerd, and @vincentkoc.
**🔒 Security/config/tooling:** Guard MCP HTTP redirects, protect global agent config defaults, and keep malformed operational limits bounded and explicit. Thanks @pgondhi987, @vincentkoc, and @joshavant.
**Channels:** WhatsApp restarts when per-account config changes, bounds background startup waits, closes failed sockets, and preserves reconnect behavior; Mattermost slash commands keep their state on `globalThis`; Feishu streaming cards preserve full merged content; voice-call tracks Twilio streams after connect; ClickClack reply tools respect `toolsAllow`. Thanks @MukundaKatta, @mcaxtr, @infoanton, @mushuiyu886, @sahibzada-allahyar, @borntobefree2-cmyk, @Takhoffman, @mmaps, @MMMMSSSS8899, @ly85206559, @sliverp, @vincentkoc, @joshavant, @LiuwqGit, @itsuzef, @wjm7220, and @donkeykong91.
**Safer channel output:** QQBot strips model reasoning and thinking scaffolding before native delivery, so users see the final answer rather than raw internal markup. Thanks @openperf, @dygg2001, @Takhoffman, and @vincentkoc.
**MCP results no longer poison sessions:** `resource_link`, audio, malformed images, and future non-text blocks are normalized before provider conversion, avoiding Anthropic 400s and broken follow-up history. Thanks @RanSHammer, @849261680, and @Takhoffman.
**Anthropic extended thinking recovers after restarts:** prompt-cache expiry and early-signature failures wait for a real message start, so they reach the existing retry path rather than ending a turn. Thanks @openperf, @MIHHHMIH, @Takhoffman, @vincentkoc, @itsuzef, and @mcaxtr.
**Parallel web search is bundled:** API-key discovery, guarded endpoint handling, cache-safe session IDs, onboarding, and documentation make it a first-class search provider.
****Improved:** Matrix and Vertex reliability:** Matrix voice notes and threaded conversations retain their context, while Google Vertex ADC models regain static catalog and runtime resolution. Thanks @849261680, @frankdierolf, and @paulogogs.
**Safer upgrades and restarts:** cron JSON stores migrate before runtime, service env placeholders **Fixed:** Masking secrets, and macOS avoids unnecessary direct-Gateway reconnect churn. Thanks @MonkeyLeeT, @sallyom, @vrurg, @wlassalle724, @jalehman, @Takhoffman, @zhangguiping-xydt, @joshavant, and @vincentkoc.
**Search/providers:** Add the Parallel bundled web-search plugin, registration contracts, onboarding integration, and guarded `api.parallel.ai/v1/search` support. Thanks @NormallyGaussian and @vincentkoc.
**Matrix/channels:** Add voice-message preflight and thread-aware read/reply behavior.
**Skills/ClawHub:** Install ClawHub skills backed by GitHub repositories through the resolved install API, download the pinned GitHub commit, keep install-policy checks, and report install telemetry after success. Thanks @Patrick-Erichsen, @vincentkoc, @itsuzef, and @mcaxtr.
**Google Chat/channels:** Add native approval card actions and click handling so Google Chat approvals use platform-native cards instead of generic message flow.
...and 3 more
🐛 Bug Fixes
**Agents:** `sessions_send` now honors an explicit `sessionKey` when stale label metadata is also present, and denied session-id sends no longer echo the resolved canonical session key. Fixes #64699; refs #74009 and #41199 Thanks @Mintalix, @RevisitMoon, @Mocha-s, @chouxiaozi1989, @sunxq1017-hash, @vincentkoc, and @joshavant.
**Channel content boundaries:** QQBot now strips reasoning/thinking tags before sending, preserving final answers while hiding internal model narration from users. Thanks @openperf, @Takhoffman, @dygg2001, and @vincentkoc.
**Agents/MCP/providers:** Coerce non-text/image MCP tool-result blocks before they reach provider converters, preserving valid images and turning richer MCP content into text instead of malformed image blocks. Thanks @RanSHammer, @849261680, @Takhoffman, @vincentkoc, and @LiuwqGit.
**Anthropic/Codex/ACP/agent recovery:** Defer Anthropic stream start events until `message_start`, strip stale compaction thinking signatures before Anthropic replay, detect unsigned thinking-only stalls, refresh prompt fences after compaction writes, reject empty completion handoffs, preserve parent streaming-off overrides/shared progress commentary, forward heartbeat metadata to context-engine hooks, and cover Codex session/thread migration edge cases. Thanks @openperf, @100yenadmin, @ooiuuii, @johnib, @Takhoffman, @MIHHHMIH, @dexiosmb, @zenglingbiao, @jalehman, @huangxun375-stack, @holgergruenhagen, @vincentkoc, @joshavant, and @ArthurusDent.
**Provider/model resolution:** Preserve Google Vertex ADC auth markers in generated catalogs, re-probe a single-provider primary after cooldown, share Codex model visibility, fail closed for unknown model auth, preserve Codex alias availability, keep unresolved profile refs unknown, and avoid resolving auth while listing models. Thanks @849261680, @paulogogs, @brtkwr, and @vincentkoc.
**Gateway/macOS/mobile:** Avoid duplicate Gateway probe warnings by identity, rate-limit node pairing requests while preserving paired-node reconnects, keep macOS node mode on a healthy direct Gateway session, keep iOS diagnostics and gateway rows reachable, and avoid Linux ARM Gradle resource tasks during Android builds. Thanks @giodl73-repo, @vrurg, @pgondhi987, @Takhoffman, @zhangguiping-xydt, @vincentkoc, @joshavant, and @shakkernerd.
TUI/chat/Workboard/auto-reply: optimistic user messages stay stable across stale history reloads, runId reassignment, and abort windows instead of disappearing, jumping, or lingering as ghost rows; Workboard stale lifecycle bulk updates no longer overwrite newer status/provenance; message-tool sends now count as delivery. Thanks @RomneyDa, @BunsDev, @kzclaw, @mcaxtr, @vincentkoc, @joshavant, and @LiuwqGit.
**Cron/update/service env:** Doctor config preflight now migrates legacy cron JSON stores into SQLite before runtime reads, service env planning skips unresolved placeholders that would mask state-dir `.env` values, and session transcript rewrites keep registry markers/discriminants consistent. Thanks @MonkeyLeeT, @sallyom, @Kvikkulf, @jalehman, @wlassalle724, @shakkernerd, and @vincentkoc.
**🔒 Security/config/tooling:** Guard MCP HTTP redirects, protect global agent config defaults, and keep malformed operational limits bounded and explicit. Thanks @pgondhi987, @vincentkoc, and @joshavant.
**Channels:** WhatsApp restarts when per-account config changes, bounds background startup waits, closes failed sockets, and preserves reconnect behavior; Mattermost slash commands keep their state on `globalThis`; Feishu streaming cards preserve full merged content; voice-call tracks Twilio streams after connect; ClickClack reply tools respect `toolsAllow`. Thanks @MukundaKatta, @mcaxtr, @infoanton, @mushuiyu886, @sahibzada-allahyar, @borntobefree2-cmyk, @Takhoffman, @mmaps, @MMMMSSSS8899, @ly85206559, @sliverp, @vincentkoc, @joshavant, @LiuwqGit, @itsuzef, @wjm7220, and @donkeykong91.
**Governed plugin and skill installs:** the old dangerous-code scanner gives way to an operator install policy with clearer doctor, CLI, ClawHub, package, archive, source, upload, and marketplace recovery paths. Thanks @joshavant and @vincentkoc.
**Safer channel delivery:** Telegram, Feishu, Discord, WhatsApp, and outbound sends now handle transcript mirroring, streamed finals, admin writeback, approval allowlists, poll modifiers, and setup state without corrupting delivery. Thanks @pgondhi987, @Petru2224, @zhangguiping-xydt, @ppmuzyk, @codezz, @takhoffman, @vincentkoc, @harjothkhara, @obviyus, @glenn-agent, @kesslerio, and @leiJack-lo.
**Steadier chat and operator UI:** visible stream text, completed sends, Workboard keyboard navigation, dialog accessibility, lazy usage views, and Android companion flows retain their state through normal interaction.
**Stricter safety checks:** config, policy, shell snapshots, exec prechecks, script limits, and Gateway startup reject malformed or unsafe input before it becomes runtime state. Thanks @RomneyDa, @giodl73-repo, @mmaps, @drobison00, @vincentkoc, and @q1387154-spec.
**More reliable Gateway and model sessions:** session locks, abandoned Codex startup, ACP handoffs, custom-provider fanout, provider aliases, prompt caching, and memory checks recover without leaving a run wedged. Thanks @RomneyDa, @takhoffman, @spencer2211, and @vincentkoc.
**🔒 Plugins/security:** Replace dangerous-code scanner enforcement with operator install policy, install-policy context, doctor checks, install/update CLI wiring, ClawHub metadata paths, and package/archive/source/upload lifecycle coverage. Thanks @joshavant and @vincentkoc.
**Policy:** Add data-handling conformance checks and reject unsupported policy keys.
**Telegram/channels:** Show commentary and reasoning in progress drafts, share progress draft compositors across channel plugins, and keep Telegram polling stop/reset boundaries cheaper and more reliable.
**UI:** Add Workboard keyboard movement controls, tighten Workboard card operations, and improve Android companion-first shell UX.
🐛 Bug Fixes
**Channels/outbound:** Keep channel sends durable when transcript mirroring fails, stop schema-padded poll modifiers from blocking normal sends, preserve WebChat `sessions_send` handoffs, preserve Discord channel-label suppression while hiding internal agent failure traces, match Discord libopus error shapes, and sanitize Discord tool progress scaffolding. Thanks @Petru2224, @codezz, @takhoffman, @harjothkhara, and @vincentkoc.
**Telegram/Feishu:** Require admin rights for Telegram target writeback, keep Telegram DM exec approval allowlists working with `ask:off`, prevent Telegram preview duplication across streaming modes, isolate verbose status after streamed finals, cancel clean restart stop timers, slow polling restart storms, and wire Feishu setup runtime setters. Thanks @pgondhi987, @zhangguiping-xydt, @ppmuzyk, @takhoffman, @vincentkoc, @obviyus, @kesslerio, @glenn-agent, and @leiJack-lo.
**Feishu:** Preserve full streaming card content by sending the merged text on each update instead of only the latest delta, so card readers see complete output when intermediate frames are missed.
**UI:** Preserve visible chat stream text, clear stale stream buffers before terminal commits, reconcile completed sends, scroll pending sends into view, harden Workboard dialog accessibility, stabilize WebChat prompt-cache affinity, overlap chat catalog startup, render chat history incrementally, lazy-load usage dashboard, and report gateway health auth diagnostics. Thanks @RomneyDa and @vincentkoc.
**CLI:** Avoid importing the runtime plugin loader only to clear in-process caches after short-lived plugin install, enable, disable, update, and uninstall commands refresh registry metadata.
**Resilient agent and Codex runs:** interrupted tool calls, stale session bindings, compaction handoffs, auth-profile failover, reasoning-tag cleanup, yielded subagents, and generated-media delivery all recover without leaving work stranded. Thanks @RomneyDa, @neeravmakwana, @joshavant, @omarshahine, @vincentkoc, @bgmbgm94, and @ksiyuna-claw.
**Reliable channel and mobile delivery:** WhatsApp, iMessage, Discord state, QQBot, and iOS Talk now preserve replies, typing, session state, QR-login recovery, and realtime connections across normal restart and transport failure paths. Thanks @mcaxtr, @omarshahine, @sliverp, @Jensenwgd, @ngutman, @vincentkoc, and @alfredjbclaw.
****Performance:** Faster Control UI chat:** startup, local drafts, incremental stream rendering, transcript caching, first connect, and post-send cleanup no longer compete with the active conversation. Thanks @vincentkoc and @sallyom.
**Governed skills and plugins:** Skill Workshop proposals, disabled-skill snapshots, support-file approvals, plugin contracts, and external package boundaries are clearer and safer for operators. Thanks @zeus1959, @sallyom, @RomneyDa, @maverikva, @vincentkoc, and @shakkernerd.
**Richer operator coordination:** Workboard goals, task-backed runs, and SQLite-backed plugin state make multi-agent work and installed-plugin discovery survive reloads cleanly.
**More dependable providers:** Google defaults, provider IDs, model catalogs, OAuth/device-code flows, media timers, and reasoning output now stay valid across hosted and local runtimes. Thanks @1052326311, @charles-openclaw, @zz327455573, @849261680, @xzh-xydt, @azgardtek, @google, @mrbrl, @nyuDSA, @vincentkoc, and @cjalden.
**Skills:** Let the `skill_workshop` agent tool apply, reject, and quarantine explicit proposals through the guarded review flow. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let pending proposals be revised in place with versioned, dated proposal frontmatter before approval.
**Skills:** Add Skill Workshop with pending proposals, CLI/Gateway review actions, rollback metadata, and the `skill_workshop` agent tool. Thanks @shakkernerd and @vincentkoc.
...and 20 more
🐛 Bug Fixes
**UI:** Render independent CLI startup metadata help snapshots concurrently to cut cold build-all metadata time.
**Plugins:** Stop timed-out package-boundary prep steps by process group so descendant TypeScript/helper processes do not survive local check cleanup.
**UI:** Serve static assets asynchronously after safe-open checks so large UI files do not block Gateway request handling.
**UI:** Forward direct wrapper SIGHUP shutdown to child processes so terminal hangups do not leave wrapped dev commands running.
**Gateway:** Return the post-expiration pending-work revision from node drains so reconnecting nodes do not observe stale queue revisions after expired items are pruned.
**Update:** Keep core updates nonblocking when missing external plugin repair downloads or soft plugin repair warnings would otherwise stall, pin post-core plugin compatibility to the downgraded core version, and still block installed active plugin payload smoke failures. Thanks @TurboTheTurtle, @Niriakot, @MukundaKatta, @giodl73-repo, @vincentkoc, and @davinci282828.
**Agents/providers:** Keep streaming tool-call argument parsing record-shaped when providers emit valid non-object JSON such as `null` or arrays.
**Talk:** Preserve explicit `null` payloads on controller-created turn and output-audio lifecycle events.
**UI:** Keep local custom provider runs from loading plugin runtime and auth alias metadata when plugins are disabled.
**Resilient agent and Codex runs:** interrupted tool calls, stale session bindings, compaction handoffs, auth-profile failover, reasoning-tag cleanup, yielded subagents, and generated-media delivery all recover without leaving work stranded. Thanks @RomneyDa, @neeravmakwana, @joshavant, @omarshahine, @vincentkoc, @bgmbgm94, and @ksiyuna-claw.
**Reliable channel and mobile delivery:** WhatsApp, iMessage, Discord state, QQBot, and iOS Talk now preserve replies, typing, session state, QR-login recovery, and realtime connections across normal restart and transport failure paths. Thanks @mcaxtr, @omarshahine, @sliverp, @Jensenwgd, @ngutman, @vincentkoc, and @alfredjbclaw.
****Performance:** Faster Control UI chat:** startup, local drafts, incremental stream rendering, transcript caching, first connect, and post-send cleanup no longer compete with the active conversation. Thanks @vincentkoc and @sallyom.
**Governed skills and plugins:** Skill Workshop proposals, disabled-skill snapshots, support-file approvals, plugin contracts, and external package boundaries are clearer and safer for operators. Thanks @zeus1959, @sallyom, @RomneyDa, @maverikva, @vincentkoc, and @shakkernerd.
**Richer operator coordination:** Workboard goals, task-backed runs, and SQLite-backed plugin state make multi-agent work and installed-plugin discovery survive reloads cleanly.
**More dependable providers:** Google defaults, provider IDs, model catalogs, OAuth/device-code flows, media timers, and reasoning output now stay valid across hosted and local runtimes. Thanks @1052326311, @charles-openclaw, @zz327455573, @849261680, @xzh-xydt, @azgardtek, @google, @mrbrl, @nyuDSA, @vincentkoc, and @cjalden.
**Skills:** Let the `skill_workshop` agent tool apply, reject, and quarantine explicit proposals through the guarded review flow. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let pending proposals be revised in place with versioned, dated proposal frontmatter before approval.
**Skills:** Add Skill Workshop with pending proposals, CLI/Gateway review actions, rollback metadata, and the `skill_workshop` agent tool. Thanks @shakkernerd and @vincentkoc.
...and 20 more
🐛 Bug Fixes
**UI:** Render independent CLI startup metadata help snapshots concurrently to cut cold build-all metadata time.
**Plugins:** Stop timed-out package-boundary prep steps by process group so descendant TypeScript/helper processes do not survive local check cleanup.
**UI:** Serve static assets asynchronously after safe-open checks so large UI files do not block Gateway request handling.
**UI:** Forward direct wrapper SIGHUP shutdown to child processes so terminal hangups do not leave wrapped dev commands running.
**Gateway:** Return the post-expiration pending-work revision from node drains so reconnecting nodes do not observe stale queue revisions after expired items are pruned.
**Update:** Keep core updates nonblocking when missing external plugin repair downloads or soft plugin repair warnings would otherwise stall, pin post-core plugin compatibility to the downgraded core version, and still block installed active plugin payload smoke failures. Thanks @TurboTheTurtle, @Niriakot, @MukundaKatta, @giodl73-repo, @vincentkoc, and @davinci282828.
**Agents/providers:** Keep streaming tool-call argument parsing record-shaped when providers emit valid non-object JSON such as `null` or arrays.
**Talk:** Preserve explicit `null` payloads on controller-created turn and output-audio lifecycle events.
**UI:** Keep local custom provider runs from loading plugin runtime and auth alias metadata when plugins are disabled.
**Resilient agent and Codex runs:** interrupted tool calls, stale session bindings, compaction handoffs, auth-profile failover, reasoning-tag cleanup, yielded subagents, and generated-media delivery all recover without leaving work stranded. Thanks @RomneyDa, @neeravmakwana, @joshavant, @omarshahine, @vincentkoc, @bgmbgm94, and @ksiyuna-claw.
**Reliable channel and mobile delivery:** WhatsApp, iMessage, Discord state, QQBot, and iOS Talk now preserve replies, typing, session state, QR-login recovery, and realtime connections across normal restart and transport failure paths. Thanks @mcaxtr, @omarshahine, @sliverp, @Jensenwgd, @ngutman, @vincentkoc, and @alfredjbclaw.
****Performance:** Faster Control UI chat:** startup, local drafts, incremental stream rendering, transcript caching, first connect, and post-send cleanup no longer compete with the active conversation. Thanks @vincentkoc and @sallyom.
**Governed skills and plugins:** Skill Workshop proposals, disabled-skill snapshots, support-file approvals, plugin contracts, and external package boundaries are clearer and safer for operators. Thanks @zeus1959, @sallyom, @RomneyDa, @maverikva, @vincentkoc, and @shakkernerd.
**Richer operator coordination:** Workboard goals, task-backed runs, and SQLite-backed plugin state make multi-agent work and installed-plugin discovery survive reloads cleanly.
**More dependable providers:** Google defaults, provider IDs, model catalogs, OAuth/device-code flows, media timers, and reasoning output now stay valid across hosted and local runtimes. Thanks @1052326311, @charles-openclaw, @zz327455573, @849261680, @xzh-xydt, @azgardtek, @google, @mrbrl, @nyuDSA, @vincentkoc, and @cjalden.
**Skills:** Let the `skill_workshop` agent tool apply, reject, and quarantine explicit proposals through the guarded review flow. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let pending proposals be revised in place with versioned, dated proposal frontmatter before approval.
**Skills:** Add Skill Workshop with pending proposals, CLI/Gateway review actions, rollback metadata, and the `skill_workshop` agent tool. Thanks @shakkernerd and @vincentkoc.
...and 20 more
🐛 Bug Fixes
**UI:** Render independent CLI startup metadata help snapshots concurrently to cut cold build-all metadata time.
**Plugins:** Stop timed-out package-boundary prep steps by process group so descendant TypeScript/helper processes do not survive local check cleanup.
**UI:** Serve static assets asynchronously after safe-open checks so large UI files do not block Gateway request handling.
**UI:** Forward direct wrapper SIGHUP shutdown to child processes so terminal hangups do not leave wrapped dev commands running.
**Gateway:** Return the post-expiration pending-work revision from node drains so reconnecting nodes do not observe stale queue revisions after expired items are pruned.
**Update:** Keep core updates nonblocking when missing external plugin repair downloads or soft plugin repair warnings would otherwise stall, pin post-core plugin compatibility to the downgraded core version, and still block installed active plugin payload smoke failures. Thanks @TurboTheTurtle, @Niriakot, @MukundaKatta, @giodl73-repo, @vincentkoc, and @davinci282828.
**Agents/providers:** Keep streaming tool-call argument parsing record-shaped when providers emit valid non-object JSON such as `null` or arrays.
**Talk:** Preserve explicit `null` payloads on controller-created turn and output-audio lifecycle events.
**UI:** Keep local custom provider runs from loading plugin runtime and auth alias metadata when plugins are disabled.
**Resilient agent and Codex runs:** interrupted tool calls, stale session bindings, compaction handoffs, auth-profile failover, reasoning-tag cleanup, yielded subagents, and generated-media delivery all recover without leaving work stranded. Thanks @RomneyDa, @neeravmakwana, @joshavant, @omarshahine, @vincentkoc, @bgmbgm94, and @ksiyuna-claw.
**Reliable channel and mobile delivery:** WhatsApp, iMessage, Discord state, QQBot, and iOS Talk now preserve replies, typing, session state, QR-login recovery, and realtime connections across normal restart and transport failure paths. Thanks @mcaxtr, @omarshahine, @sliverp, @Jensenwgd, @ngutman, @vincentkoc, and @alfredjbclaw.
****Performance:** Faster Control UI chat:** startup, local drafts, incremental stream rendering, transcript caching, first connect, and post-send cleanup no longer compete with the active conversation. Thanks @vincentkoc and @sallyom.
**Governed skills and plugins:** Skill Workshop proposals, disabled-skill snapshots, support-file approvals, plugin contracts, and external package boundaries are clearer and safer for operators. Thanks @zeus1959, @sallyom, @RomneyDa, @maverikva, @vincentkoc, and @shakkernerd.
**Richer operator coordination:** Workboard goals, task-backed runs, and SQLite-backed plugin state make multi-agent work and installed-plugin discovery survive reloads cleanly.
**More dependable providers:** Google defaults, provider IDs, model catalogs, OAuth/device-code flows, media timers, and reasoning output now stay valid across hosted and local runtimes. Thanks @1052326311, @charles-openclaw, @zz327455573, @849261680, @xzh-xydt, @azgardtek, @google, @mrbrl, @nyuDSA, @vincentkoc, and @cjalden.
**Skills:** Let the `skill_workshop` agent tool apply, reject, and quarantine explicit proposals through the guarded review flow. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards. Thanks @shakkernerd and @vincentkoc.
**Skills:** Let pending proposals be revised in place with versioned, dated proposal frontmatter before approval.
**Skills:** Add Skill Workshop with pending proposals, CLI/Gateway review actions, rollback metadata, and the `skill_workshop` agent tool. Thanks @shakkernerd and @vincentkoc.
...and 20 more
🐛 Bug Fixes
**UI:** Render independent CLI startup metadata help snapshots concurrently to cut cold build-all metadata time.
**Plugins:** Stop timed-out package-boundary prep steps by process group so descendant TypeScript/helper processes do not survive local check cleanup.
**UI:** Serve static assets asynchronously after safe-open checks so large UI files do not block Gateway request handling.
**UI:** Forward direct wrapper SIGHUP shutdown to child processes so terminal hangups do not leave wrapped dev commands running.
**Gateway:** Return the post-expiration pending-work revision from node drains so reconnecting nodes do not observe stale queue revisions after expired items are pruned.
**Update:** Keep core updates nonblocking when missing external plugin repair downloads or soft plugin repair warnings would otherwise stall, pin post-core plugin compatibility to the downgraded core version, and still block installed active plugin payload smoke failures. Thanks @TurboTheTurtle, @Niriakot, @MukundaKatta, @giodl73-repo, @vincentkoc, and @davinci282828.
**Agents/providers:** Keep streaming tool-call argument parsing record-shaped when providers emit valid non-object JSON such as `null` or arrays.
**Talk:** Preserve explicit `null` payloads on controller-created turn and output-audio lifecycle events.
**UI:** Keep local custom provider runs from loading plugin runtime and auth alias metadata when plugins are disabled.
****Performance:** Faster Control UI chat:** transcript rendering, draft persistence, pending-send paint, scrolling, and first-message work no longer block one another while a conversation is active.
**More reliable agent recovery:** fallback auth state, legacy Codex app-server auth, stale bootstrap history, and streamed reasoning cleanup recover cleanly instead of carrying stale turn state forward. Thanks @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Clearer cron and channel history:** external `sessions_send` routes, single-job history, and channel reply state retain the right destination and job context through restarts. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
****Improved:** Operator coordination:** core session goals and durable plugin-install state make long-running work easier to inspect and resume after reloads.
**Safer mobile reconnects:** iOS realtime sessions retain their WebSocket continuation instead of failing a healthy Talk connection, with hosted push/realtime paths kept aligned. Thanks @ngutman and @vincentkoc.
Control UI now streams stable Markdown blocks and records more chat responsiveness signals to guide first-reply performance work.
🐛 Bug Fixes
**UI:** Keep first Control UI sends responsive, retain pending sends while history catches up, cache transcript renders, and avoid draft persistence or navigation churn from blocking the active conversation.
**🔒 Agents/Codex/auth:** Repair automatic fallback state, accept supported legacy Codex app-server auth, remove stale bootstrap history, strip streamed reasoning tags, and validate shell snapshots against trusted environment data. Thanks @RomneyDa, @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Cron/channels:** Preserve external `sessions_send` routes, include the job name in single-job history, keep Mattermost attachments on the upload path, and bound Telegram/installer/Parallels proof cleanup. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
**PR #87469** feat: add core session goals.
**PR #88129** fix: keep live OpenClaw session locks during cleanup.
**PR #88133** **Fixed:** Heartbeat default run timeout. Related #87438.
**PR #88132** fix: move compaction planning off the event loop. Related #86358.
**PR #88137** **Fixed:** Ci mainline checks.
**PR #88140** perf: centralize skills indexing and visibility.
**PR #82326** **Plugin manifest contract for SecretRef** is now supported as a provider integrations. Thanks @sallyom and @joshavant.
****Performance:** Faster Control UI chat:** transcript rendering, draft persistence, pending-send paint, scrolling, and first-message work no longer block one another while a conversation is active.
**More reliable agent recovery:** fallback auth state, legacy Codex app-server auth, stale bootstrap history, and streamed reasoning cleanup recover cleanly instead of carrying stale turn state forward. Thanks @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Clearer cron and channel history:** external `sessions_send` routes, single-job history, and channel reply state retain the right destination and job context through restarts. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
****Improved:** Operator coordination:** core session goals and durable plugin-install state make long-running work easier to inspect and resume after reloads.
**Safer mobile reconnects:** iOS realtime sessions retain their WebSocket continuation instead of failing a healthy Talk connection, with hosted push/realtime paths kept aligned. Thanks @ngutman and @vincentkoc.
Control UI now streams stable Markdown blocks and records more chat responsiveness signals to guide first-reply performance work.
🐛 Bug Fixes
**UI:** Keep first Control UI sends responsive, retain pending sends while history catches up, cache transcript renders, and avoid draft persistence or navigation churn from blocking the active conversation.
**🔒 Agents/Codex/auth:** Repair automatic fallback state, accept supported legacy Codex app-server auth, remove stale bootstrap history, strip streamed reasoning tags, and validate shell snapshots against trusted environment data. Thanks @RomneyDa, @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Cron/channels:** Preserve external `sessions_send` routes, include the job name in single-job history, keep Mattermost attachments on the upload path, and bound Telegram/installer/Parallels proof cleanup. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
**PR #87469** feat: add core session goals.
**PR #88129** fix: keep live OpenClaw session locks during cleanup.
**PR #88133** **Fixed:** Heartbeat default run timeout. Related #87438.
**PR #88132** fix: move compaction planning off the event loop. Related #86358.
**PR #88137** **Fixed:** Ci mainline checks.
**PR #88140** perf: centralize skills indexing and visibility.
**PR #82326** **Plugin manifest contract for SecretRef** is now supported as a provider integrations. Thanks @sallyom and @joshavant.
****Performance:** Faster Control UI chat:** transcript rendering, draft persistence, pending-send paint, scrolling, and first-message work no longer block one another while a conversation is active.
**More reliable agent recovery:** fallback auth state, legacy Codex app-server auth, stale bootstrap history, and streamed reasoning cleanup recover cleanly instead of carrying stale turn state forward. Thanks @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Clearer cron and channel history:** external `sessions_send` routes, single-job history, and channel reply state retain the right destination and job context through restarts. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
****Improved:** Operator coordination:** core session goals and durable plugin-install state make long-running work easier to inspect and resume after reloads.
**Safer mobile reconnects:** iOS realtime sessions retain their WebSocket continuation instead of failing a healthy Talk connection, with hosted push/realtime paths kept aligned. Thanks @ngutman and @vincentkoc.
Control UI now streams stable Markdown blocks and records more chat responsiveness signals to guide first-reply performance work.
🐛 Bug Fixes
**UI:** Keep first Control UI sends responsive, retain pending sends while history catches up, cache transcript renders, and avoid draft persistence or navigation churn from blocking the active conversation.
**🔒 Agents/Codex/auth:** Repair automatic fallback state, accept supported legacy Codex app-server auth, remove stale bootstrap history, strip streamed reasoning tags, and validate shell snapshots against trusted environment data. Thanks @RomneyDa, @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Cron/channels:** Preserve external `sessions_send` routes, include the job name in single-job history, keep Mattermost attachments on the upload path, and bound Telegram/installer/Parallels proof cleanup. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
**PR #87469** feat: add core session goals.
**PR #88129** fix: keep live OpenClaw session locks during cleanup.
**PR #88133** **Fixed:** Heartbeat default run timeout. Related #87438.
**PR #88132** fix: move compaction planning off the event loop. Related #86358.
**PR #88137** **Fixed:** Ci mainline checks.
**PR #88140** perf: centralize skills indexing and visibility.
**PR #82326** **Plugin manifest contract for SecretRef** is now supported as a provider integrations. Thanks @sallyom and @joshavant.
****Performance:** Faster Control UI chat:** transcript rendering, draft persistence, pending-send paint, scrolling, and first-message work no longer block one another while a conversation is active.
**More reliable agent recovery:** fallback auth state, legacy Codex app-server auth, stale bootstrap history, and streamed reasoning cleanup recover cleanly instead of carrying stale turn state forward. Thanks @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Clearer cron and channel history:** external `sessions_send` routes, single-job history, and channel reply state retain the right destination and job context through restarts. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
****Improved:** Operator coordination:** core session goals and durable plugin-install state make long-running work easier to inspect and resume after reloads.
**Safer mobile reconnects:** iOS realtime sessions retain their WebSocket continuation instead of failing a healthy Talk connection, with hosted push/realtime paths kept aligned. Thanks @ngutman and @vincentkoc.
Control UI now streams stable Markdown blocks and records more chat responsiveness signals to guide first-reply performance work.
🐛 Bug Fixes
**UI:** Keep first Control UI sends responsive, retain pending sends while history catches up, cache transcript renders, and avoid draft persistence or navigation churn from blocking the active conversation.
**🔒 Agents/Codex/auth:** Repair automatic fallback state, accept supported legacy Codex app-server auth, remove stale bootstrap history, strip streamed reasoning tags, and validate shell snapshots against trusted environment data. Thanks @RomneyDa, @neeravmakwana, @vincentkoc, @bgmbgm94, and @kdonthar.
**Cron/channels:** Preserve external `sessions_send` routes, include the job name in single-job history, keep Mattermost attachments on the upload path, and bound Telegram/installer/Parallels proof cleanup. Thanks @kip-claw, @MonkeyLeeT, @vincentkoc, and @Lvan185.
**PR #87469** feat: add core session goals.
**PR #88129** fix: keep live OpenClaw session locks during cleanup.
**PR #88133** **Fixed:** Heartbeat default run timeout. Related #87438.
**PR #88132** fix: move compaction planning off the event loop. Related #86358.
**PR #88137** **Fixed:** Ci mainline checks.
**PR #88140** perf: centralize skills indexing and visibility.
**PR #82326** **Plugin manifest contract for SecretRef** is now supported as a provider integrations. Thanks @sallyom and @joshavant.
Agents and CLI-backed runtimes recover more cleanly from interrupted tool calls, stale session bindings, compaction handoffs, and media delivery retries.
Channels and mobile delivery are steadier across Telegram, WhatsApp, iMessage, Slack, Discord, Microsoft Teams, Google Chat, Google Meet, and iOS realtime Talk.
Provider and plugin requests now bound more timers, retries, OAuth/device-code lifetimes, media downloads, local service probes, and generated-content polling paths before they can hang a run.
Skills, session metadata, gateway runtime state, plugin metadata, and store writes do less repeated work on hot paths while keeping config and dispatch behavior stable.
Workboard, SecretRef plugin manifests, hosted iOS push relay, and external Copilot/Tokenjuice packaging **Broader orchestration,** is now supported as a provider, and plugin delivery surfaces.
Release, CI, Docker, E2E, and diagnostics lanes now cap more logs, response bodies, readiness probes, artifact checks, and status polling so failures report bounded proof instead of stalling.
**Skills:** Let the `skill_research` agent tool apply, reject, and quarantine explicit Skill Workshop proposals through the guarded proposal lifecycle.
**Skills:** Let Skill Workshop proposals carry approved support files under standard skill folders, with scanner, hash, and rollback safeguards.
**Skills:** Let pending Skill Workshop proposals be revised in place with versioned, dated proposal frontmatter before approval.
**Skills:** Add Skill Workshop proposals with pending `PROPOSAL.md` drafts, CLI/Gateway review actions, rollback metadata, and the `skill_research` agent tool.
...and 8 more
🐛 Bug Fixes
**CLI:** Keep `plugins list --json` on the snapshot-only path so plugin sweeps avoid loading the full runtime status graph.
**Plugins:** Make PixVerse external-plugin ClawHub metadata explicit and keep it out of bundled dist builds.
**Agent and Codex runtime recovery is steadier:** Subagents keep cwd/workspace separation, hook context stays prompt-local, session locks release on timeout abort while live OpenClaw locks survive cleanup, stale restart continuations are avoided, and Codex app-server/helper failures no longer tear down shared runtime state.
Channel delivery and session identity got safer across outbound plugin hooks, Matrix room ids, iMessage reactions/approvals, Slack final replies, Discord recovered tool warnings, runtime-config message actions, WhatsApp profile auth roots, Telegram polling, and Microsoft Teams service URL trust checks.
**Mobile and chat surfaces got a broader refresh:** The iOS Pro UI, hosted push relay default, realtime Talk tab playback, Gateway chat transport, onboarding, Talk permissions, WebChat reconnect delivery, and session picker behavior now preserve more state across reconnects and empty searches. Thanks @ngutman and @BunsDev.
Browser, channel, and automation inputs are stricter: Browser tool timeouts, viewport/tab indices, Gateway ports, cron retry handling, Discord component ids, schema array refs, Telegram callback pages, and channel progress callbacks now reject malformed values earlier and preserve the intended delivery context.
Provider, media, and document coverage expands with Claude Opus 4.8, Fal Krea image schemas, NVIDIA featured models, MiniMax streaming music responses, encrypted PDF extraction, voice model catalogs, GitHub Copilot agent runtime support, and a Codex Supervisor plugin path for delegated Codex workflows.
CLI, auth, doctor, and provider paths fail **Performance:** Faster and recover more clearly: malformed numeric/version options are rejected, workspace dotenv provider credentials are ignored, heartbeat defaults, OAuth/token lifetimes, and local service startup requests are bounded, agent auth health labels are clearer, legacy `api_key` auth profiles migrate to canonical form, and restart guidance is actionable. Thanks @vincentkoc and @giodl73-repo.
Plugin and Gateway hot paths do less repeated work while preserving cache correctness for install records, config JSON parsing, tool search catalogs, session stores, manifest model rows, auto-enabled plugin config, browser tokens, viewer assets, and release-split external plugin packages.
Release, QA, and E2E validation now bound more log, artifact, harness, and cross-OS waits so failing lanes produce proof instead of hanging or false-greening.
**Status:** Show active subagent details in status output.
**Diffs:** Split the default language pack and expand default Diffs language coverage while keeping the host floor aligned.
...and 11 more
🐛 Bug Fixes
**Agents:** Fall back to local config pruning when the optional `agents delete` Gateway probe cannot authenticate, so offline installs can still delete agents without removing shared workspaces.
**Agents/Codex:** Keep spawned agent cwd/workspace state separated, forward ACP spawn attachments, keep hook context prompt-local, release session locks on timeout abort and runtime teardown without deleting live OpenClaw-owned locks during cleanup, avoid session event queue self-wait, clean up exec abort listeners, stream assistant deltas incrementally, recover raw missing-thread compaction failures, preserve rotated compaction session identity, keep compaction-timeout snapshots continuable, preserve shared app-server state across startup or helper failures, keep native hook relay alive across restarts and prune stale bridge files, close native hook relay replacement races, keep Claude live tool progress visible for watchdog recovery, suppress abandoned requester completion handoff, route workspace memory through tools, resolve Codex runtime models first, report quarantined dynamic tools, format `skills` command output, bind node auto-review to prepared plans, retry Claude CLI transcript probes, and bound compaction/steering retries. Thanks @mbelinky, @Alix-007, @luoyanglang, @yetval, @sjf, @joshavant, @benjamin1492, @c19354837, @fuller-stack-dev, @pfrederiksen, and @dodge1218.
**Codex Supervisor:** Keep real-home app-server MCP session listing on the loaded state path, bound stored history scans, and close WebSocket probes cleanly.
**Channels:** Thread canonical session keys into outbound hooks, preserve Matrix room-id case, keep fallback tool warnings mention-inert, retain delivered Slack final replies during late cleanup, continue iMessage polling after denied reactions, suppress duplicate native exec approvals, resolve Gateway message actions against the active runtime config, preserve Telegram SecretRef prompt config and polling keepalives, preserve WhatsApp profile auth roots, QR display, document filenames, and plugin hook config, suppress Discord recovered tool warnings, preserve the Discord voice outbound helper, cap Discord/Signal/Zalo channel request and container timeouts, and block untrusted Teams service URLs while keeping TeamsSDK patterns aligned. Thanks @zeroaltitude, @lukeboyett, @jarvis-mns1, @xiaotian, @funmerlin, @joshavant, @eleqtrizit, @heyitsaamir, @amittell, @lidge-jun, @liorb-mountapps, @masatohoshino, @bladin, and @giodl73-repo.
**🔒 CLI/auth/doctor/providers:** Reject malformed numeric/timeout/subcommand-version inputs, ignore workspace dotenv provider credentials, wait for respawn child shutdown, bound heartbeat defaults plus Codex, GitHub Copilot, OpenAI, Anthropic, Google, Feishu, LM Studio, MiniMax, Xiaomi TTS, and local-provider OAuth/token/model requests, harden Codex auth probes, label auth health by agent, preserve explicit agentRuntime pins during Codex model migration, warm provider auth off the main thread, honor Codex response timeouts, stop migrating current Claude Haiku 4.5 profiles to Sonnet, bound local service startup, resolve GPT-5.5 without cached catalog, migrate legacy memory auto-provider config, rewrite non-canonical `api_key` auth profiles, and make doctor restart follow-ups actionable. Thanks @Patrick-Erichsen, @samzong, @giodl73-repo, @alkor2000, @mmaps, @nxmxbbd, and @vincentkoc.
**🔒 Gateway/security/session state:** Expire browser tokens after auth rotation, scope assistant idempotency dedupe, drain probe client closes, avoid stale restart continuation reuse, preserve retry-after fallbacks and stale rate-limit cooldown probes, bound webchat image and artifact transcript scans, include seconds in inbound metadata timestamps, clear completed session active runs, clear stale chat stream buffers, and evict current plugin-state namespaces at row caps. Thanks @joshavant and @litang9.
**Config/parsing/network:** Reject partial numeric parsing, parse provider/Discord retry headers and dates strictly, honor IPv6 and bare IPv6 `no_proxy` entries, preserve empty plugin allowlists, canonicalize secret target array indexes, and reject malformed media content lengths, inspected TCP ports, marketplace content lengths, cron epochs, sandbox stat fields, unsafe duration values, empty config path segments, noncanonical schema array refs, unsafe Telegram callback pages, and invalid Teams attachment-fetch DNS targets.
**Agent and Codex runtime recovery is steadier:** Subagents keep cwd/workspace separation, hook context stays prompt-local, session locks release on timeout abort, stale restart continuations are avoided, and Codex app-server/helper failures no longer tear down shared runtime state.
Channel delivery and session identity got safer across outbound plugin hooks, Matrix room ids, iMessage reactions/approvals, Slack final replies, Discord recovered tool warnings, WhatsApp profile auth roots, Telegram polling, and Microsoft Teams service URL trust checks.
**Mobile and chat surfaces got a broader refresh:** The iOS Pro UI, hosted push relay default, realtime Talk tab playback, Gateway chat transport, onboarding, Talk permissions, WebChat reconnect delivery, and session picker behavior now preserve more state across reconnects and empty searches.
Browser, channel, and automation inputs are stricter: Browser tool timeouts, viewport/tab indices, Gateway ports, cron retry handling, Discord component ids, schema array refs, Telegram callback pages, and channel progress callbacks now reject malformed values earlier and preserve the intended delivery context.
Provider, media, and document coverage expands with Claude Opus 4.8, Fal Krea image schemas, NVIDIA featured models, MiniMax streaming music responses, encrypted PDF extraction, voice model catalogs, GitHub Copilot agent runtime support, and a Codex Supervisor plugin path for delegated Codex workflows.
CLI, auth, doctor, and provider paths fail **Performance:** Faster and recover more clearly: malformed numeric/version options are rejected, workspace dotenv provider credentials are ignored, OAuth/token lifetimes and local service startup requests are bounded, agent auth health labels are clearer, legacy `api_key` auth profiles migrate to canonical form, and restart guidance is actionable. Thanks @vincentkoc and @giodl73-repo.
Plugin and Gateway hot paths do less repeated work while preserving cache correctness for install records, config JSON parsing, tool search catalogs, session stores, manifest model rows, auto-enabled plugin config, browser tokens, viewer assets, and release-split external plugin packages.
Release, QA, and E2E validation now bound more log, artifact, harness, and cross-OS waits so failing lanes produce proof instead of hanging or false-greening.
**Status:** Show active subagent details in status output.
**Diffs:** Split the default language pack and expand default Diffs language coverage while keeping the host floor aligned.
...and 11 more
🐛 Bug Fixes
**Agents:** Fall back to local config pruning when the optional `agents delete` Gateway probe cannot authenticate, so offline installs can still delete agents without removing shared workspaces.
**Agents/Codex:** Keep spawned agent cwd/workspace state separated, forward ACP spawn attachments, keep hook context prompt-local, release session locks on timeout abort and runtime teardown, avoid session event queue self-wait, clean up exec abort listeners, stream assistant deltas incrementally, recover raw missing-thread compaction failures, preserve rotated compaction session identity, keep compaction-timeout snapshots continuable, preserve shared app-server state across startup or helper failures, keep native hook relay alive across restarts and prune stale bridge files, close native hook relay replacement races, keep Claude live tool progress visible for watchdog recovery, suppress abandoned requester completion handoff, route workspace memory through tools, resolve Codex runtime models first, report quarantined dynamic tools, format `skills` command output, bind node auto-review to prepared plans, retry Claude CLI transcript probes, and bound compaction/steering retries. Thanks @mbelinky, @Alix-007, @luoyanglang, @yetval, @sjf, @joshavant, and @benjamin1492.
**Codex Supervisor:** Keep real-home app-server MCP session listing on the loaded state path, bound stored history scans, and close WebSocket probes cleanly.
**Channels:** Thread canonical session keys into outbound hooks, preserve Matrix room-id case, keep fallback tool warnings mention-inert, retain delivered Slack final replies during late cleanup, continue iMessage polling after denied reactions, suppress duplicate native exec approvals, preserve Telegram SecretRef prompt config and polling keepalives, preserve WhatsApp profile auth roots, QR display, document filenames, and plugin hook config, suppress Discord recovered tool warnings, preserve the Discord voice outbound helper, cap Discord/Signal/Zalo channel request and container timeouts, and block untrusted Teams service URLs while keeping TeamsSDK patterns aligned. Thanks @zeroaltitude, @lukeboyett, @xiaotian, @eleqtrizit, @heyitsaamir, @amittell, @liorb-mountapps, @masatohoshino, @bladin, and @giodl73-repo.
**🔒 CLI/auth/doctor/providers:** Reject malformed numeric/timeout/subcommand-version inputs, ignore workspace dotenv provider credentials, wait for respawn child shutdown, bound Codex, GitHub Copilot, OpenAI, Anthropic, Google, Feishu, LM Studio, MiniMax, Xiaomi TTS, and local-provider OAuth/token/model requests, harden Codex auth probes, label auth health by agent, preserve explicit agentRuntime pins during Codex model migration, warm provider auth off the main thread, honor Codex response timeouts, stop migrating current Claude Haiku 4.5 profiles to Sonnet, bound local service startup, resolve GPT-5.5 without cached catalog, migrate legacy memory auto-provider config, rewrite non-canonical `api_key` auth profiles, and make doctor restart follow-ups actionable. Thanks @Patrick-Erichsen, @samzong, @giodl73-repo, @alkor2000, @mmaps, @nxmxbbd, and @vincentkoc.
**🔒 Gateway/security/session state:** Expire browser tokens after auth rotation, scope assistant idempotency dedupe, drain probe client closes, avoid stale restart continuation reuse, preserve retry-after fallbacks and stale rate-limit cooldown probes, bound webchat image and artifact transcript scans, include seconds in inbound metadata timestamps, clear completed session active runs, clear stale chat stream buffers, and evict current plugin-state namespaces at row caps. Thanks @joshavant and @litang9.
**Config/parsing/network:** Reject partial numeric parsing, parse provider/Discord retry headers and dates strictly, honor IPv6 and bare IPv6 `no_proxy` entries, preserve empty plugin allowlists, canonicalize secret target array indexes, and reject malformed media content lengths, inspected TCP ports, marketplace content lengths, cron epochs, sandbox stat fields, unsafe duration values, empty config path segments, noncanonical schema array refs, unsafe Telegram callback pages, and invalid Teams attachment-fetch DNS targets.
**Agent and Codex runtime recovery is steadier:** Subagents keep cwd/workspace separation, hook context stays prompt-local, session locks release on timeout abort, stale restart continuations are avoided, and Codex app-server/helper failures no longer tear down shared runtime state.
Channel delivery and session identity got safer across outbound plugin hooks, Matrix room ids, iMessage reactions/approvals, Slack final replies, Discord recovered tool warnings, WhatsApp profile auth roots, Telegram polling, and Microsoft Teams service URL trust checks.
**Mobile and chat surfaces got a broader refresh:** The iOS Pro UI, Gateway chat transport, onboarding, Talk permissions, WebChat reconnect delivery, and session picker behavior now preserve more state across reconnects and empty searches.
Browser, channel, and automation inputs are stricter: Browser tool timeouts, viewport/tab indices, Gateway ports, cron retry handling, Discord component ids, schema array refs, Telegram callback pages, and channel progress callbacks now reject malformed values earlier and preserve the intended delivery context.
Provider, media, and document coverage expands with Claude Opus 4.8, Fal Krea image schemas, NVIDIA featured models, MiniMax streaming music responses, encrypted PDF extraction, voice model catalogs, GitHub Copilot agent runtime support, and a Codex Supervisor plugin path for delegated Codex workflows.
CLI, auth, doctor, and provider paths fail **Performance:** Faster and recover more clearly: malformed numeric/version options are rejected, workspace dotenv provider credentials are ignored, OAuth and local service startup requests are bounded, legacy `api_key` auth profiles migrate to canonical form, and restart guidance is actionable.
Plugin and Gateway hot paths do less repeated work while preserving cache correctness for install records, config JSON parsing, tool search catalogs, session stores, manifest model rows, auto-enabled plugin config, browser tokens, and viewer assets.
Release, QA, and E2E validation now bound more log, artifact, harness, and cross-OS waits so failing lanes produce proof instead of hanging or false-greening.
**Status:** Show active subagent details in status output.
**Diffs:** Split the default language pack and expand default Diffs language coverage while keeping the host floor aligned.
...and 9 more
🐛 Bug Fixes
**Agents:** Fall back to local config pruning when the optional `agents delete` Gateway probe cannot authenticate, so offline installs can still delete agents without removing shared workspaces.
**Cron/automation:** Retry recurring jobs after transient model rate limits before waiting for the next scheduled slot, and preflight model fallbacks before skipping scheduled work.
**Agent and Codex runtime recovery is steadier:** Subagents keep cwd/workspace separation, hook context stays prompt-local, session locks release on timeout abort, stale restart continuations are avoided, and Codex app-server/helper failures no longer tear down shared runtime state.
Channel delivery and session identity got safer across outbound plugin hooks, Matrix room ids, iMessage reactions/approvals, Slack final replies, Discord recovered tool warnings, WhatsApp profile auth roots, Telegram polling, and Microsoft Teams service URL trust checks.
**Mobile and chat surfaces got a broader refresh:** The iOS Pro UI, Gateway chat transport, onboarding, Talk permissions, WebChat reconnect delivery, and session picker behavior now preserve more state across reconnects and empty searches.
Browser, channel, and automation inputs are stricter: Browser tool timeouts, viewport/tab indices, Gateway ports, cron retry handling, Discord component ids, schema array refs, Telegram callback pages, and channel progress callbacks now reject malformed values earlier and preserve the intended delivery context.
Provider, media, and document coverage expands with Claude Opus 4.8, Fal Krea image schemas, NVIDIA featured models, MiniMax streaming music responses, encrypted PDF extraction, voice model catalogs, GitHub Copilot agent runtime support, and a Codex Supervisor plugin path for delegated Codex workflows.
CLI, auth, doctor, and provider paths fail **Performance:** Faster and recover more clearly: malformed numeric/version options are rejected, workspace dotenv provider credentials are ignored, OAuth and local service startup requests are bounded, legacy `api_key` auth profiles migrate to canonical form, and restart guidance is actionable.
Plugin and Gateway hot paths do less repeated work while preserving cache correctness for install records, config JSON parsing, tool search catalogs, session stores, manifest model rows, auto-enabled plugin config, browser tokens, and viewer assets.
Release, QA, and E2E validation now bound more log, artifact, harness, and cross-OS waits so failing lanes produce proof instead of hanging or false-greening.
**Status:** Show active subagent details in status output.
**Diffs:** Split the default language pack and expand default Diffs language coverage while keeping the host floor aligned.
...and 9 more
🐛 Bug Fixes
**Agents:** Fall back to local config pruning when the optional `agents delete` Gateway probe cannot authenticate, so offline installs can still delete agents without removing shared workspaces.
**Cron/automation:** Retry recurring jobs after transient model rate limits before waiting for the next scheduled slot, and preflight model fallbacks before skipping scheduled work.
**Agent and Codex runtime recovery is steadier:** Subagents keep cwd/workspace separation, hook context stays prompt-local, session locks release on timeout abort, stale restart continuations are avoided, and Codex app-server/helper failures no longer tear down shared runtime state.
Channel delivery and session identity got safer across outbound plugin hooks, Matrix room ids, iMessage reactions/approvals, Slack final replies, Discord recovered tool warnings, and Microsoft Teams service URL trust checks.
**Mobile and chat surfaces got a broader refresh:** The iOS Pro UI, Gateway chat transport, onboarding, Talk permissions, WebChat reconnect delivery, and session picker behavior now preserve more state across reconnects and empty searches.
CLI, auth, doctor, and provider paths fail **Performance:** Faster and recover more clearly: malformed numeric/version options are rejected, OAuth and local service startup requests are bounded, legacy `api_key` auth profiles migrate to canonical form, and restart guidance is actionable.
Plugin and Gateway hot paths do less repeated work while preserving cache correctness for install records, config JSON parsing, tool search catalogs, session stores, manifest model rows, auto-enabled plugin config, browser tokens, and viewer assets.
Release, QA, and E2E validation now bound more log, artifact, harness, and cross-OS waits so failing lanes produce proof instead of hanging or false-greening.
**Status:** Show active subagent details in status output.
**Diffs:** Split the default language pack and expand default Diffs language coverage while keeping the host floor aligned.
**ClawHub:** Add plugin display names plus skill verification and trust surfaces. Thanks @thewilloftheshadow and @Patrick-Erichsen.
**iOS:** Refresh the dev app with Pro Command, Chat, Agents, and Settings tabs wired to gateway sessions, diagnostics, chat, and realtime Talk.
...and 2 more
🐛 Bug Fixes
**Agents:** Fall back to local config pruning when the optional `agents delete` Gateway probe cannot authenticate, so offline installs can still delete agents without removing shared workspaces.
**Agents/Codex:** Keep spawned agent cwd/workspace state separated, keep hook context prompt-local, release session locks on timeout abort, avoid session event queue self-wait, preserve shared app-server state across startup or helper failures, keep native hook relay alive across restarts, route workspace memory through tools, resolve Codex runtime models first, report quarantined dynamic tools, format `skills` command output, and bound compaction/steering retries. Thanks @mbelinky, @Alix-007, @luoyanglang, @yetval, and @sjf.
**Channels:** Thread canonical session keys into outbound hooks, preserve Matrix room-id case, keep fallback tool warnings mention-inert, retain delivered Slack final replies during late cleanup, continue iMessage polling after denied reactions, suppress duplicate native exec approvals, preserve Telegram SecretRef prompt config, suppress Discord recovered tool warnings, and block untrusted Teams service URLs. Thanks @zeroaltitude, @lukeboyett, @xiaotian, and @eleqtrizit.
**🔒 CLI/auth/doctor/providers:** Reject malformed numeric/timeout/subcommand-version inputs, wait for respawn child shutdown, bound Codex and GitHub Copilot OAuth/token requests, warm provider auth off the main thread, honor Codex response timeouts, bound local service startup, resolve GPT-5.5 without cached catalog, migrate legacy memory auto-provider config, rewrite non-canonical `api_key` auth profiles, and make doctor restart follow-ups actionable. Thanks @Patrick-Erichsen, @samzong, @giodl73-repo, and @alkor2000.
**🔒 Gateway/security/session state:** Expire browser tokens after auth rotation, scope assistant idempotency dedupe, drain probe client closes, avoid stale restart continuation reuse, preserve retry-after fallbacks, bound webchat image and artifact transcript scans, include seconds in inbound metadata timestamps, and evict current plugin-state namespaces at row caps.
**Config/parsing/network:** Reject partial numeric parsing, parse provider/Discord retry headers and dates strictly, honor IPv6 and bare IPv6 `no_proxy` entries, canonicalize secret target array indexes, and reject malformed media content lengths, inspected TCP ports, marketplace content lengths, cron epochs, and sandbox stat fields.
**🔒 Stronger security and content boundaries:** Group prompt text is kept out of the system prompt, repeated-dot hostnames are normalized, side-effecting command wrappers and unsafe Node runtime env overrides are blocked, no-auth Tailscale exposure is rejected, and node/device-role approvals now require admin authority. Thanks @eleqtrizit and @pgondhi987.
More reliable Codex app-server runs: Codex runtime models **Fixed:** First, workspace memory is routed through tools, shared app-server clients survive startup and spawned-helper failures, native hook relay generations survive restarts and rotate on fresh fallbacks, and false runtime live switches are avoided.
**Faster Gateway and reply paths:** Session reads, plugin metadata fingerprints, auth env snapshots, auto-enabled plugin config, tool-search catalogs, and stable metadata caches do less hot-path rediscovery while visible replies no longer inherit hidden cleanup timeouts.
**Better provider and model coverage:** OpenAI-compatible embedding providers are core, DeepInfra catalog browsing loads the full credential-aware model set, Pixverse adds video generation and API region selection, VLLM thinking params are wired, Claude CLI OAuth overlays load for PI auth profiles, and bare direct Anthropic model ids work. Thanks @dutifulbob, @ats3v, and @joshavant.
**Channel delivery is steadier:** Telegram `sendMessage` actions use durable outbound delivery, iMessage suppresses duplicate native exec approval prompts and sends, Slack keeps delivered final replies during late cleanup, Matrix mention previews/finals are stricter, QQBot fallback approval buttons honor slash-command auth, Discord guild requester checks are tighter, recovered Discord tool-warning artifacts stay out of successful replies, and Google Chat stops thread sends in DMs. Thanks @mbelinky and @eleqtrizit.
Release, package, and CI proof paths are harder to wedge: npm/package inventory honors dist exclusions, shrinkwrap override pins merge correctly, Docker runtime workspace templates are packaged and smoked, release postpublish checks are stricter, beta smoke rejects empty runs, and E2E log/probe waits are bounded.
**Memory:** Add a core OpenAI-compatible embedding provider for local and hosted OpenAI-style endpoints, with config, doctor, and docs support.
**Plugin SDK:** Mark memory-specific embedding provider registration as deprecated compatibility and surface non-bundled usage in plugin compatibility diagnostics.
**Providers:** Add the Pixverse video generation provider, API region selection, docs, and external plugin packaging support.
**DeepInfra:** Load the full model catalog when users browse models during onboarding, preserve configured API-key catalogs, refresh media/video defaults, and keep pricing/default model metadata aligned.
...and 4 more
🐛 Bug Fixes
**🔒 Security/content boundaries:** Route untrusted group prompt metadata outside system prompts, normalize repeated trailing hostname dots, block side-effecting command wrappers, reject unsafe Node runtime env overrides, reject no-auth Tailscale exposure, block untrusted Microsoft Teams service URLs, enforce `/allowlist configWrites` origin policy, gate QQBot fallback approval buttons, and require admin for node/device-role approvals. Thanks @eleqtrizit and @pgondhi987.
**Codex:** Resolve Codex runtime models before generic routing, route workspace memory through tools, preserve shared app-server clients after startup and spawned-helper failures, preserve native hook relay generations across restarts and fresh fallbacks, keep raw reasoning/source-reply guards intact, report quarantined dynamic tools, keep the attempt watchdog armed for queued terminal turns, and route Codex OAuth compaction through OpenAI-Codex.
**Agents/runtime:** Avoid session event queue self-waits, bound compaction wake and steering retries, preserve grace for pending error diagnostics, avoid false Codex runtime live switches, avoid stale restart continuation reuse, preserve session fallback errors, suppress duplicate Claude CLI skill prompts, keep runtime context before active user turns, strip stale Anthropic thinking, quarantine unsupported tool schemas, recover completed write timeouts safely, release retained session write locks on timeout abort, and validate forced plugin harness support before pinning. Thanks @luoyanglang, @cathrynlavery, and @openperf.
**Reply/session delivery:** Keep visible turn admission unbounded, keep visible fallback delivery on latest targets, preserve bridge hook context, classify direct fallback targets by channel grammar, report approval resolutions in bridge mode, and avoid stale source-reply artifacts.
**Channels:** Make Telegram `sendMessage` action replies durable and preserve SecretRef prompt config, suppress duplicate iMessage native exec approval prompts and sends, keep iMessage approval polling alive after denied reactions, keep Slack delivered final replies during late cleanup, keep Matrix mention previews/finals mention-inert and normally delivered, ignore filename-embedded Matrix IDs, suppress recovered Discord tool-warning artifacts from successful replies, suppress Google Chat thread sends in DMs, and harden Discord guild requester checks.
**Memory:** Salvage QMD search JSON after nonzero exits and keep workspace memory routing through the Codex tool path where possible.
**Providers/models:** Forward cached token usage in OpenAI-compatible chat completions, load Claude CLI OAuth overlays for PI auth profiles, send bare direct Anthropic model ids, wire configured VLLM thinking params, honor OpenAI-compatible cache retention, normalize OpenAI Responses replay tool ids, resolve OpenAI `gpt-5.5` without a cached catalog, preserve `retry-after` fallback handling, bound GitHub Copilot auth requests, and load DeepInfra custom/live catalogs consistently. Thanks @caz0075, @joshavant, and @ats3v.
**Gateway/performance:** Borrow read-only session metadata and active session working stores, cache current/stable plugin metadata fingerprints, cache auto-enabled plugin config, slim metadata identity caches, trust current metadata lifecycle caches, stabilize isolated cron prompt-cache affinity, persist model auth profile suffixes, drain probe client closes, expire browser tokens after auth rotation, and keep default status fast paths bounded.
**CLI:** Reject loose or malformed numeric options for gateway timeouts, model limits, directory limits, message options, webhooks, and partial values; respect subcommand version options; route generated/root/plugin help targets correctly; keep skills JSON output flushing naturally; and keep plugin descriptor loading quiet in root help.
**Plugin state/tool search:** Evict the current namespace when plugin rows hit caps, reuse unchanged tool-search catalogs, align the release catalog reuse wrapper, and keep fallback tool warnings mention-inert.
**🔒 Stronger security and content boundaries:** Group prompt text is kept out of the system prompt, repeated-dot hostnames are normalized, side-effecting command wrappers and unsafe Node runtime env overrides are blocked, no-auth Tailscale exposure is rejected, and node/device-role approvals now require admin authority. Thanks @eleqtrizit and @pgondhi987.
More reliable Codex app-server runs: Codex runtime models **Fixed:** First, workspace memory is routed through tools, shared app-server clients survive startup and spawned-helper failures, native hook relay generations survive restarts and rotate on fresh fallbacks, and false runtime live switches are avoided.
**Faster Gateway and reply paths:** Session reads, plugin metadata fingerprints, auth env snapshots, auto-enabled plugin config, tool-search catalogs, and stable metadata caches do less hot-path rediscovery while visible replies no longer inherit hidden cleanup timeouts.
**Better provider and model coverage:** OpenAI-compatible embedding providers are core, DeepInfra catalog browsing loads the full credential-aware model set, Pixverse adds video generation and API region selection, VLLM thinking params are wired, Claude CLI OAuth overlays load for PI auth profiles, and bare direct Anthropic model ids work. Thanks @dutifulbob, @ats3v, and @joshavant.
**Channel delivery is steadier:** Telegram `sendMessage` actions use durable outbound delivery, iMessage suppresses duplicate native exec approval prompts and sends, Slack keeps delivered final replies during late cleanup, Matrix mention previews/finals are stricter, QQBot fallback approval buttons honor slash-command auth, Discord guild requester checks are tighter, recovered Discord tool-warning artifacts stay out of successful replies, and Google Chat stops thread sends in DMs. Thanks @mbelinky and @eleqtrizit.
Release, package, and CI proof paths are harder to wedge: npm/package inventory honors dist exclusions, shrinkwrap override pins merge correctly, Docker runtime workspace templates are packaged and smoked, release postpublish checks are stricter, beta smoke rejects empty runs, and E2E log/probe waits are bounded.
**Memory:** Add a core OpenAI-compatible embedding provider for local and hosted OpenAI-style endpoints, with config, doctor, and docs support.
**Plugin SDK:** Mark memory-specific embedding provider registration as deprecated compatibility and surface non-bundled usage in plugin compatibility diagnostics.
**Providers:** Add the Pixverse video generation provider, API region selection, docs, and external plugin packaging support.
**DeepInfra:** Load the full model catalog when users browse models during onboarding, preserve configured API-key catalogs, refresh media/video defaults, and keep pricing/default model metadata aligned.
...and 4 more
🐛 Bug Fixes
**🔒 Security/content boundaries:** Route untrusted group prompt metadata outside system prompts, normalize repeated trailing hostname dots, block side-effecting command wrappers, reject unsafe Node runtime env overrides, reject no-auth Tailscale exposure, block untrusted Microsoft Teams service URLs, enforce `/allowlist configWrites` origin policy, gate QQBot fallback approval buttons, and require admin for node/device-role approvals. Thanks @eleqtrizit and @pgondhi987.
**Codex:** Resolve Codex runtime models before generic routing, route workspace memory through tools, preserve shared app-server clients after startup and spawned-helper failures, preserve native hook relay generations across restarts and fresh fallbacks, keep raw reasoning/source-reply guards intact, report quarantined dynamic tools, keep the attempt watchdog armed for queued terminal turns, and route Codex OAuth compaction through OpenAI-Codex.
**Agents/runtime:** Avoid session event queue self-waits, bound compaction wake and steering retries, preserve grace for pending error diagnostics, avoid false Codex runtime live switches, avoid stale restart continuation reuse, preserve session fallback errors, suppress duplicate Claude CLI skill prompts, keep runtime context before active user turns, strip stale Anthropic thinking, quarantine unsupported tool schemas, recover completed write timeouts safely, release retained session write locks on timeout abort, and validate forced plugin harness support before pinning. Thanks @luoyanglang, @cathrynlavery, and @openperf.
**Reply/session delivery:** Keep visible turn admission unbounded, keep visible fallback delivery on latest targets, preserve bridge hook context, classify direct fallback targets by channel grammar, report approval resolutions in bridge mode, and avoid stale source-reply artifacts.
**Channels:** Make Telegram `sendMessage` action replies durable and preserve SecretRef prompt config, suppress duplicate iMessage native exec approval prompts and sends, keep iMessage approval polling alive after denied reactions, keep Slack delivered final replies during late cleanup, keep Matrix mention previews/finals mention-inert and normally delivered, ignore filename-embedded Matrix IDs, suppress recovered Discord tool-warning artifacts from successful replies, suppress Google Chat thread sends in DMs, and harden Discord guild requester checks.
**Memory:** Salvage QMD search JSON after nonzero exits and keep workspace memory routing through the Codex tool path where possible.
**Providers/models:** Forward cached token usage in OpenAI-compatible chat completions, load Claude CLI OAuth overlays for PI auth profiles, send bare direct Anthropic model ids, wire configured VLLM thinking params, honor OpenAI-compatible cache retention, normalize OpenAI Responses replay tool ids, resolve OpenAI `gpt-5.5` without a cached catalog, preserve `retry-after` fallback handling, bound GitHub Copilot auth requests, and load DeepInfra custom/live catalogs consistently. Thanks @caz0075, @joshavant, and @ats3v.
**Gateway/performance:** Borrow read-only session metadata and active session working stores, cache current/stable plugin metadata fingerprints, cache auto-enabled plugin config, slim metadata identity caches, trust current metadata lifecycle caches, stabilize isolated cron prompt-cache affinity, persist model auth profile suffixes, drain probe client closes, expire browser tokens after auth rotation, and keep default status fast paths bounded.
**CLI:** Reject loose or malformed numeric options for gateway timeouts, model limits, directory limits, message options, webhooks, and partial values; respect subcommand version options; route generated/root/plugin help targets correctly; keep skills JSON output flushing naturally; and keep plugin descriptor loading quiet in root help.
**Plugin state/tool search:** Evict the current namespace when plugin rows hit caps, reuse unchanged tool-search catalogs, align the release catalog reuse wrapper, and keep fallback tool warnings mention-inert.
**Faster Gateway and replies:** Startup avoids repeated plugin, channel, session, usage-cost, warning, scheduled-service, and filesystem scans; visible replies separate user-facing sends from slower follow-up work; Gateway runtime/session caches churn less under load.
**Transcripts are core:** Transcript-backed meeting summaries, source-provider chunks, cleaned user turns, media provenance, Codex mirrors, WebChat replies, and CLI/TUI replay now use one more reliable transcript path.
More channels are production-ready: Telegram keeps typing/progress context and forum topics, iMessage handles attachment roots, remote media staging, and duplicate local Messages sources, WhatsApp restores group/media behavior, Discord improves voice playback and model picking, and Signal/iMessage/WhatsApp get reaction approvals.
**Better voice and Talk:** Realtime Talk runs can be inspected, steered, cancelled, or followed up from Web UI and Discord voice; wake-name handling is more tolerant without letting ambient speech trigger agents.
**Safer content boundaries:** Browser snapshot reads honor SSRF policy, system-event text cannot spoof nested prompt markers, fetched file text is wrapped as external content, ClickClack inbound sender allowlists run before agent dispatch, stale device tokens are rejected, and serialized tool-call text is scrubbed from replies.
Providers, Codex, and local models are steadier: named auth profiles, OpenAI sampling params, Codex app-server resume/timeout/usage-limit recovery, dynamic tool-schema guards, xAI usage-limit surfacing, Ollama top-p normalization, and local approval resolution reduce provider-specific dead ends.
**Transcripts:** Add core transcript capture and source-provider support for transcript-backed meeting summaries, including the renamed Transcripts docs, CLI surface, source-provider chunks, and cleaned user-turn persistence.
**🔒 Auth:** Add named model login profiles and supported credential migration for Hermes, OpenCode, and Codex auth profiles, with explicit opt-out and non-interactive controls.
...and 21 more
🐛 Bug Fixes
**🔒 Memory/security:** Reject prompt-like text submitted through the explicit `memory_store` tool before embedding or storage, matching the existing auto-capture prompt-injection filter.
**🔒 Gateway/security:** Enable the default auth rate limiter for remote non-browser and HTTP gateway auth failures when `gateway.auth.rateLimit` is unset, while preserving the loopback exemption.
**🔒 Security/content boundaries:** Validate Browser snapshot tab URLs against SSRF policy before ChromeMCP or direct CDP reads, sanitize queued system-event text so untrusted plugin/channel labels cannot spoof nested prompt markers, wrap fetched file text and metadata as external content, apply ClickClack `allowFrom` sender allowlists before agent dispatch, reject RPCs from invalidated device-token clients during rotation, require staged sandbox media refs, and scrub serialized tool-call text from replies. Thanks @zsxsoft, @ttzero25, and @mmaps.
**Transcripts/user turns:** Persist CLI, WebChat, media, follow-up, hook, and Codex-mirror user turns to the admitted session target; keep cleaned transcript text, inline image routing, provenance metadata, replay hooks, and fallback paths idempotent when runtimes fail or restart.
**UI:** Queue busy TUI prompts instead of dropping them, preserve the configured default model during onboarding, show failed tool results as errors, show config-open failures in Control UI, keep status JSON plugin scans healthy, preserve xAI usage-limit errors locally, and expose explicit fast-mode/systemd state.
**CLI:** Preserve plugin LLM command auth, bind native plugin command dispatch to the host agent's LLM auth, keep `onDiagnosticEvent` exports discoverable through `Function.name`, stabilize diagnostic event root aliases, correlate pathless read diagnostics, suppress transient runner failures in channel command paths, and repair local approval resolution.
**Codex/providers:** Keep WebChat delivery hints out of user prompts, avoid false queued-terminal idle timeouts, share the native hook relay registry, quarantine unsupported dynamic tool schemas, preserve Claude resumed-session system prompts, normalize greedy Ollama `top_p`, preserve per-agent thinking defaults for ingress runs, and avoid native compaction takeover on budget-triggered Codex turns.
**Gateway/perf/release:** Reuse startup-warning metadata and prepared auth stores, avoid cloning live-switch and lifecycle session caches on read paths, defer warning and scheduled-service fallback imports, trim Gateway session/startup/runtime CPU churn, skip duplicate turn session touches, stop chat timeout fallback cascades, drop stale subagent announce history, bound benchmark/watch/kitchen-sink teardown waits, bound macOS/package/onboarding/plugin smoke commands, bound install finalization probes, resolve Parallels npm-update commands from guest `PATH`, and bootstrap raw AWS macOS Node/pnpm commands through `/usr/bin/env`.
**Reply/source delivery:** Keep TUI, Control UI, media, TTS, transcript, and Codex source-reply finals live without duplicate terminal events or stale replay artifacts.
**Faster Gateway and replies:** Startup avoids repeated plugin, channel, session, usage-cost, warning, scheduled-service, and filesystem scans; visible replies separate user-facing sends from slower follow-up work; Gateway runtime/session caches churn less under load.
**Transcripts are core:** Transcript-backed meeting summaries, source-provider chunks, cleaned user turns, media provenance, Codex mirrors, WebChat replies, and CLI/TUI replay now use one more reliable transcript path.
More channels are production-ready: Telegram keeps typing/progress context and forum topics, iMessage handles attachment roots, remote media staging, and duplicate local Messages sources, WhatsApp restores group/media behavior, Discord improves voice playback and model picking, and Signal/iMessage/WhatsApp get reaction approvals.
**Better voice and Talk:** Realtime Talk runs can be inspected, steered, cancelled, or followed up from Web UI and Discord voice; wake-name handling is more tolerant without letting ambient speech trigger agents.
**Safer content boundaries:** Browser snapshot reads honor SSRF policy, system-event text cannot spoof nested prompt markers, fetched file text is wrapped as external content, ClickClack inbound sender allowlists run before agent dispatch, stale device tokens are rejected, and serialized tool-call text is scrubbed from replies.
Providers, Codex, and local models are steadier: named auth profiles, OpenAI sampling params, Codex app-server resume/timeout/usage-limit recovery, dynamic tool-schema guards, xAI usage-limit surfacing, Ollama top-p normalization, and local approval resolution reduce provider-specific dead ends.
**Transcripts:** Add core transcript capture and source-provider support for transcript-backed meeting summaries, including the renamed Transcripts docs, CLI surface, source-provider chunks, and cleaned user-turn persistence.
**🔒 Auth:** Add named model login profiles and supported credential migration for Hermes, OpenCode, and Codex auth profiles, with explicit opt-out and non-interactive controls.
...and 21 more
🐛 Bug Fixes
**🔒 Memory/security:** Reject prompt-like text submitted through the explicit `memory_store` tool before embedding or storage, matching the existing auto-capture prompt-injection filter.
**🔒 Gateway/security:** Enable the default auth rate limiter for remote non-browser and HTTP gateway auth failures when `gateway.auth.rateLimit` is unset, while preserving the loopback exemption.
**🔒 Security/content boundaries:** Validate Browser snapshot tab URLs against SSRF policy before ChromeMCP or direct CDP reads, sanitize queued system-event text so untrusted plugin/channel labels cannot spoof nested prompt markers, wrap fetched file text and metadata as external content, apply ClickClack `allowFrom` sender allowlists before agent dispatch, reject RPCs from invalidated device-token clients during rotation, require staged sandbox media refs, and scrub serialized tool-call text from replies. Thanks @zsxsoft, @ttzero25, and @mmaps.
**Transcripts/user turns:** Persist CLI, WebChat, media, follow-up, hook, and Codex-mirror user turns to the admitted session target; keep cleaned transcript text, inline image routing, provenance metadata, replay hooks, and fallback paths idempotent when runtimes fail or restart.
**UI:** Queue busy TUI prompts instead of dropping them, preserve the configured default model during onboarding, show failed tool results as errors, show config-open failures in Control UI, keep status JSON plugin scans healthy, preserve xAI usage-limit errors locally, and expose explicit fast-mode/systemd state.
**CLI:** Preserve plugin LLM command auth, bind native plugin command dispatch to the host agent's LLM auth, keep `onDiagnosticEvent` exports discoverable through `Function.name`, stabilize diagnostic event root aliases, correlate pathless read diagnostics, suppress transient runner failures in channel command paths, and repair local approval resolution.
**Codex/providers:** Keep WebChat delivery hints out of user prompts, avoid false queued-terminal idle timeouts, share the native hook relay registry, quarantine unsupported dynamic tool schemas, preserve Claude resumed-session system prompts, normalize greedy Ollama `top_p`, preserve per-agent thinking defaults for ingress runs, and avoid native compaction takeover on budget-triggered Codex turns.
**Gateway/perf/release:** Reuse startup-warning metadata and prepared auth stores, avoid cloning live-switch and lifecycle session caches on read paths, defer warning and scheduled-service fallback imports, trim Gateway session/startup/runtime CPU churn, skip duplicate turn session touches, stop chat timeout fallback cascades, drop stale subagent announce history, bound benchmark/watch/kitchen-sink teardown waits, bound macOS/package/onboarding/plugin smoke commands, bound install finalization probes, resolve Parallels npm-update commands from guest `PATH`, and bootstrap raw AWS macOS Node/pnpm commands through `/usr/bin/env`.
**Reply/source delivery:** Keep TUI, Control UI, media, TTS, transcript, and Codex source-reply finals live without duplicate terminal events or stale replay artifacts.
**Faster replies and startup:** Visible reply delivery now separates user-facing sends from slower follow-up work, command/model/plugin metadata is reused on hot paths, and Gateway startup avoids repeated plugin, channel, session, usage-cost, and filesystem scans.
**Better voice and Talk:** Realtime Talk runs can be inspected, steered, cancelled, or followed up from Web UI and Discord voice; wake-name handling is more tolerant without letting ambient speech trigger agents.
More channels are production-ready: Telegram keeps typing/progress context and forum topics, iMessage handles attachment roots and duplicate local Messages sources, WhatsApp restores group/media behavior, Discord improves voice playback and model picking, and Signal/iMessage get reaction approvals.
**Safer agents:** Codex app-server auth, compaction, source replies, sandbox path handling, and usage-limit recovery are more robust; OpenAI-compatible providers avoid empty-tool and malformed payload failures.
**More reliable replay and installs:** Legacy tool results, subagent spawn payloads, stale lock ownership, Windows stack-heavy startup, macOS restart validation, and Docker package preparation all fail less surprisingly.
**New observability:** Activity tab, gateway secret-prep traces, tool/model stream progress, OpenTelemetry LLM spans, release performance evidence, and richer missing telemetry signals make failures easier to inspect.
**Transcripts:** Add core transcript capture and source-provider support for transcript-backed meeting summaries, including the renamed Transcripts docs and CLI surface.
**🔒 Auth:** Add named model login profiles and supported credential migration for Hermes, OpenCode, and Codex auth profiles, with explicit opt-out and non-interactive controls.
**Diagnostics:** Trace gateway secret preparation, classify skill/tool usage, surface model stream progress, add OpenTelemetry LLM content spans, and expose alertable telemetry for blocked tools, failover, stale sessions, liveness, oversized payloads, and webhook ingress.
**Reply/source delivery:** Keep TUI, Control UI, media, TTS, transcript, and Codex source-reply finals live without duplicate terminal events or stale replay artifacts.
**Agents/replay:** Repair legacy tool results before replay, preserve `sessions_spawn` transcript payloads, restore current guard checks, stage sandboxed workspace media, and keep duplicate transcripts tool display metadata from reappearing. Thanks @martingarramon, @vincentkoc, and @joshavant.
**Codex:** Project newer OpenClaw chat history into resumed app-server threads and keep Codex turn timeouts inside the Codex runtime boundary so timeouts do not poison shared app-server clients or fall through to unrelated provider fallback. Thanks @TurboTheTurtle and @pashpashpash.
**Config/doctor/update:** Narrow profiled tool-section doctor repair, keep runtime-injected legacy web-search provider config out of user-authored config validation, and keep prerelease tags excluded from stable updater resolution. Thanks @joshavant, @luoyanglang, and @stevenepalmer.
**CLI:** Add a Windows-only stack-size respawn for stack-heavy startup paths, default CLI logs to local timestamps, and validate timeout/banner TTY state more strictly. Thanks @giodl73-repo and @vincentkoc.
**🔒 Locking/security:** Require owner identity proof before stale plugin lock removal, memoize session lock owner arguments, and avoid writing default exec approval stores unless policy state actually changed. Thanks @Alix-007 and @vincentkoc.
**Install/release:** Bound Docker package build, inventory, pack, and tarball preparation with process-group timeouts; pin shrinkwrap patch drift to the pnpm lock; harden macOS restart and dSYM packaging; and run release Docker/live timeout wrappers in the foreground so child processes cannot wedge gates.
**Telegram/network:** Treat `ENETDOWN` as a transient pre-connect network failure so Telegram sends, gateway unhandled-rejection handling, and cron network retries follow the same recovery path as sibling network outages.
**Telegram:** Preserve inbound text entities, overlapping DM replies, account topic cache sidecars, outbound reply context, targeted bot-command mentions, durable group retry targets, forum topic names, and native progress callbacks. Thanks @SebTardif, @luoyanglang, and @neeravmakwana.
**Installer:** Let the local-prefix CLI installer use Alpine's `apk` Node.js, npm, and Git packages on musl Linux instead of downloading glibc Node tarballs that fail `node:sqlite`.
**Checks:** Prefilter tracked conflict-marker scans so changed checks avoid reading every repository file on clean runs.
**Plugins:** Allow linked local plugin paths to probe TypeScript source entries without requiring compiled package output, restoring source-checkout plugin development on native Windows.
**CLI:** Route source-checkout build output to stderr before launching OpenClaw commands so stale local builds do not corrupt `--json` stdout.
**Installer:** Install Node.js through `apk` on Alpine Linux instead of falling through to the NodeSource package-manager path.
**Agents/perf:** Cache manifest-backed CLI provider descriptors and fallback provider resolution so model fallback retries avoid repeated bundled provider runtime scans while still invalidating across plugin reloads.
**Installer:** Detect musl Linux shells such as Alpine as Linux instead of rejecting them before npm install.
**Windows:** Run direct Node package scripts with env overrides through a cross-platform launcher so gateway, TUI, Docker-all, generated-module formatting, and optional Discord native opus installer entrypoints work on native Windows.
**Tests:** Run Vitest import timing entrypoints through a Node wrapper so native Windows package scripts can collect import diagnostics.
**UI:** Split large build-time runtime dependencies into stable chunks so Linux/Docker install and package builds stay below the app chunk warning threshold.
**iMessage:** Support thumb-approval reactions — `👍` (Like tapback) resolves an approval as `allow-once` and `👎` resolves as `deny`, with the explicit-approver allowlist read from `channels.imessage.allowFrom`; `allow-always` stays on the manual `/approve allow-always` text fallback. Mirrors the WhatsApp behavior from #85477.
**Gateway/perf:** Reuse process-stable channel catalog reads, avoid repeated bundled-channel boundary checks, and rotate gateway watch CPU profiles so benchmark runs do not accumulate unbounded artifacts.
**Gateway/perf:** Cache stable install-record, channel-catalog, bundled-channel, and Telegram session-store metadata during process-local hot paths to reduce repeated JSON and manifest reads.
**Gateway/perf:** Reuse immutable plugin metadata snapshots across startup, config, model, channel, setup, and secret metadata readers so hot paths avoid repeated plugin file stats and manifest registry reloads.
**Talk/realtime:** Let WebUI and Discord voice callers ask for active OpenClaw run status, cancel, steer, or queue follow-up work while a consult is still running.
**Discord/voice:** Add realtime wake-name gating with agent-name defaults and raise profile bootstrap context budget for longer `USER.md`/`SOUL.md` files.
**Gateway/perf:** Lazy-load startup-idle plugin work, core gateway method handlers, and the embedded ACPX runtime so Gateway health and ready signals no longer wait on unused handler trees or ACPX probes.
**Gateway/perf:** Cache plugin SDK public-surface alias maps and skip irrelevant macOS Linuxbrew PATH probes so Gateway startup avoids repeated filesystem walks and slow missing-directory stats.
**Image tool:** Add adaptive model-aware image compression with an `agents.defaults.imageQuality` preference for choosing token-efficient, balanced, or high-detail media handling.
**Meeting Notes:** Add a source-only external meeting-notes plugin and SDK source-provider contract outside the core npm package, with auto-start capture config, manual transcript imports, read-only `openclaw meeting-notes` CLI access, and Discord voice as the first live source.
...and 47 more
🐛 Bug Fixes
**Gateway/update:** Avoid fetching unrelated tags during dev-channel git updates so moved release tags do not block branch-based updates.
**CLI:** Suppress the expected future-config warning while an old update parent hands off to the freshly installed post-core process.
**MiniMax:** Store OAuth token expiry as an absolute millisecond timestamp so OAuth profiles no longer appear expired on every request.
**Agents/Anthropic:** Strip missing or blank thinking signatures for signed-thinking providers even when recovery supplies a narrow replay policy without signature preservation. Fixes #84430.
**Agents/channels:** Send a visible notice when an aborted main session cannot be resumed after restart, including Telegram group targets.
**Discord/voice:** Serialize overlapping voice joins, retry aborted startup readiness within the configured timeout, upgrade meeting-notes-only sessions to realtime when the normal follow join arrives, detach promoted meeting-notes ownership without leaving voice, and include `OpenClaw` in default realtime wake names.
**Gateway/restart:** Honor the configured restart drain budget for embedded runs and avoid spending the deferral timeout twice after forced restart timeouts.
**Gateway/boot:** Run `BOOT.md` startup checks in an isolated boot session so gateway restarts do not overwrite the agent's main session mapping.
**Meeting Notes:** Include a speaker-labeled transcript section in generated summaries so Discord group voice captures show who said each captured utterance.
**Discord/voice:** Recover stale realtime playback state when Discord stream-close/player-idle events do not arrive, and keep generated runtime plugin aliases available after postbuild rewrites.
**Gateway/perf:** Reuse process-stable channel catalog reads, avoid repeated bundled-channel boundary checks, and rotate gateway watch CPU profiles so benchmark runs do not accumulate unbounded artifacts.
**Gateway/perf:** Cache stable install-record, channel-catalog, bundled-channel, and Telegram session-store metadata during process-local hot paths to reduce repeated JSON and manifest reads.
**Gateway/perf:** Reuse immutable plugin metadata snapshots across startup, config, model, channel, setup, and secret metadata readers so hot paths avoid repeated plugin file stats and manifest registry reloads.
**Talk/realtime:** Let WebUI and Discord voice callers ask for active OpenClaw run status, cancel, steer, or queue follow-up work while a consult is still running.
**Discord/voice:** Add realtime wake-name gating with agent-name defaults and raise profile bootstrap context budget for longer `USER.md`/`SOUL.md` files.
**Gateway/perf:** Lazy-load startup-idle plugin work, core gateway method handlers, and the embedded ACPX runtime so Gateway health and ready signals no longer wait on unused handler trees or ACPX probes.
**Gateway/perf:** Cache plugin SDK public-surface alias maps and skip irrelevant macOS Linuxbrew PATH probes so Gateway startup avoids repeated filesystem walks and slow missing-directory stats.
**Image tool:** Add adaptive model-aware image compression with an `agents.defaults.imageQuality` preference for choosing token-efficient, balanced, or high-detail media handling.
**Meeting Notes:** Add a source-only external meeting-notes plugin and SDK source-provider contract outside the core npm package, with auto-start capture config, manual transcript imports, read-only `openclaw meeting-notes` CLI access, and Discord voice as the first live source.
**Meeting Notes/Discord:** Release channel account startup before meeting-notes auto-capture, wait for the Discord voice manager during gateway boot, and stop plugin services before channel shutdown so voice capture state remains available during startup and cleanup.
...and 46 more
🐛 Bug Fixes
**Gateway/update:** Avoid fetching unrelated tags during dev-channel git updates so moved release tags do not block branch-based updates.
**MiniMax:** Store OAuth token expiry as an absolute millisecond timestamp so OAuth profiles no longer appear expired on every request.
**Agents/Anthropic:** Strip missing or blank thinking signatures for signed-thinking providers even when recovery supplies a narrow replay policy without signature preservation. Fixes #84430.
**Agents/channels:** Send a visible notice when an aborted main session cannot be resumed after restart, including Telegram group targets.
**Discord/voice:** Serialize overlapping voice joins, retry aborted startup readiness within the configured timeout, upgrade meeting-notes-only sessions to realtime when the normal follow join arrives, detach promoted meeting-notes ownership without leaving voice, and include `OpenClaw` in default realtime wake names.
**Gateway/restart:** Honor the configured restart drain budget for embedded runs and avoid spending the deferral timeout twice after forced restart timeouts.
**Gateway/boot:** Run `BOOT.md` startup checks in an isolated boot session so gateway restarts do not overwrite the agent's main session mapping.
**Meeting Notes:** Include a speaker-labeled transcript section in generated summaries so Discord group voice captures show who said each captured utterance.
**Discord/voice:** Recover stale realtime playback state when Discord stream-close/player-idle events do not arrive, and keep generated runtime plugin aliases available after postbuild rewrites.
**Discord/voice:** Keep realtime playback running when meeting notes attaches to an existing voice session or a realtime consult starts, and route realtime user transcripts into meeting notes.
**Gateway/perf:** Reuse process-stable channel catalog reads, avoid repeated bundled-channel boundary checks, and rotate gateway watch CPU profiles so benchmark runs do not accumulate unbounded artifacts.
**Gateway/perf:** Reuse immutable plugin metadata snapshots across startup, config, model, channel, setup, and secret metadata readers so hot paths avoid repeated plugin file stats and manifest registry reloads.
**Gateway/perf:** Lazy-load startup-idle plugin work, core gateway method handlers, and the embedded ACPX runtime so Gateway health and ready signals no longer wait on unused handler trees or ACPX probes.
**Gateway/perf:** Cache plugin SDK public-surface alias maps and skip irrelevant macOS Linuxbrew PATH probes so Gateway startup avoids repeated filesystem walks and slow missing-directory stats.
**Meeting Notes:** Add a source-only external meeting-notes plugin and SDK source-provider contract outside the core npm package, with auto-start capture config, manual transcript imports, read-only `openclaw meeting-notes` CLI access, and Discord voice as the first live source.
**Docs/channels/config:** Add Signal `configPath`, Telegram wildcard topic defaults, local-time backup archive names, Termux home fallback, include-path validation, secret-scanner-safe placeholder guidance, Gemini CLI/Antigravity media guidance, and macOS VM auto-login guidance. Thanks @NorseGaud, @yudistiraashadi, @huangqian8, @VibhorGautam, @maweibin, @tianxingleo, @IgnacioPro, and @xzcxzcyy-claw.
**Docs:** Clarify model-usage portability, Codex migration prerequisites, status bootstrap wording, thread-bound subagent limits, hook ownership, and config-preserving safety guidance. Thanks @aniruddhaadak80, @leno23, @TomDjerry, @matthewxmurphy, @vincentkoc, and @stablegenius49.
**Docs:** Clarify README onboarding and Gateway startup paths, WhatsApp QR/408 recovery, cron output language prompts, skill advanced features, gateway upstream 403 troubleshooting, and plugin fallback override guidance. Thanks @deepujain, @Zacxxx, @Jah-yee, @neyric, @usimic, @Renu-Cybe, @BigUncle, and @SeashoreShi.
**Docs:** Clarify context-pruning ratio bounds, local dashboard recovery, CLI env markers, remote onboarding token behavior, and Peekaboo Bridge permissions for subprocess agents. Thanks @ayesha-aziz123, @dishraters, @hougangdev, and @brandonlipman.
**UI:** Summarize internal message-tool source replies so tool cards no longer duplicate the visible reply body.
**Gateway:** Preserve deferred lifecycle-error cleanup across later non-terminal events so provider timeouts can persist failed session state instead of leaving sessions stuck running. (#85256, fixes #63819).
**Agents/subagents:** Report tool-only child progress during timeout summaries instead of showing no visible output.
**Telegram/ACP:** Preserve explicit `:topic:` conversation suffixes when inbound ACP targets do not carry a separate thread id.
**Browser/proxy:** Bypass the managed proxy for the exact local managed Chrome CDP readiness and DevTools WebSocket endpoints, so `openclaw browser start` works when the operator proxy blocks loopback egress.
**Ollama:** Bypass the managed proxy for configured local embedding origins while keeping SSRF guardrails on unconfigured targets.
**OpenAI/images:** Route Codex API-key image generation through the native OpenAI Images API instead of the Codex OAuth streaming backend, avoiding 401s from valid API keys.
**Agents/OpenAI completions:** Omit empty tool payload fields for proxy-like OpenAI-compatible endpoints so strict vLLM-style servers accept tool-free turns.
**Checks/Windows:** Route full `pnpm check` stage commands through the managed child runner so Windows avoids Node shell-argv deprecation warnings there too.
**Checks/Windows:** Run managed child commands through explicit `cmd.exe` wrapping instead of Node shell mode with argv, avoiding Node 24 subprocess deprecation warnings during changed checks.
**Docs:** Consolidate GLM under Z.AI, add the Upstash Box install guide and Gateway exposure runbook, clarify MEDIA directives, Copilot and Voyage setup, config path quoting, real behavior proof, and memory-file write guidance. Thanks @BobDu, @alitariksahin, @Jefsky, @musaabhasan, @OmerZeyveli, @leno23, @WuKongAI-CMU, @luoyanglang, and @majin1102.
**Docs:** Clarify media provider credentials, Codex/OpenClaw code-mode boundaries, Slack and Telegram ack reactions, Feishu dynamic agents, secrets plaintext boundaries, memory guidance, and Chinese glossary terms. Thanks @nielskaspers, @cosmopolitan033, @drclaw-iq, @alexgduarte, @zccyman, @chengoak, and @cassthebandit.
**Packaging:** Exclude documentation images and assets from the npm tarball, reducing published package size without affecting runtime docs search or CLI behavior.
**Media understanding:** Stop auto-probing Gemini CLI and use Antigravity CLI only as a lower-priority image/video fallback after configured provider APIs.
**Agents/subagents:** Limit default sub-agent bootstrap context to `AGENTS.md` and `TOOLS.md`, keeping persona, identity, user, memory, heartbeat, and setup files out of delegated workers by default.
...and 29 more
🐛 Bug Fixes
**Agents/tools:** Honor configured custom provider API keys when deciding whether media, image-generation, video-generation, music-generation, and PDF tools are available.
**StepFun:** Stop advertising stale generic API key auth choices so onboarding only offers runtime-backed Standard and Step Plan choices.
**Diagnostics:** Keep OpenTelemetry log bodies behind explicit content capture and scrub scoped agent-session keys from OpenTelemetry and Prometheus labels while preserving bounded queue-lane prefixes.
**Windows installer:** Fail Git checkout installs when `pnpm install` or `pnpm build` fails instead of writing a wrapper to a missing CLI build.
**Sessions:** Surface previous-transcript archive failures during `/new` rotation so disk rename errors are logged instead of silently hiding stranded transcript files. Fixes #81984. (#85586, from #82081).
**UI:** Mirror internal-ui message-tool replies into final chat output so message-tool-only agents remain visible in `openclaw tui`. Fixes #85538.
**Agents:** Keep parallel OpenAI-compatible tool-call deltas in separate argument buffers so interleaved tool calls no longer corrupt streamed arguments.
**Memory/doctor:** Report missing or unusable QMD workspace directories as workspace failures instead of generic binary failures.
**Debug proxy:** Record CONNECT client-socket errors and destroy the paired upstream socket so abrupt client disconnects no longer leak tunnel resources.
**Diffs:** Continue hydrating later diff cards when one card fails so a single broken card no longer blanks the whole diff viewer.
**Exec approvals:** Remove the old `cat SKILL.md && printf ... && ` allowlist compatibility path so skill files must be loaded with the read tool and only the real skill executable is auto-allowed.
**Discord:** Let voice sessions follow configured Discord users into voice channels, with allowed-channel checks, multi-user handoff, bounded reconciliation, and DAVE recovery preservation.
**Discord/voice:** Include bounded `IDENTITY.md`, `USER.md`, and `SOUL.md` profile context in realtime voice session instructions by default, with `voice.realtime.bootstrapContextFiles: []` available to disable it.
**Dependencies:** Bump the bundled Codex harness to `@openai/codex` `0.132.0` and refresh the app-server model-list docs for the new catalog.
**CLI:** Add the bundled Policy plugin for policy-backed channel conformance checks, doctor lint findings, and opt-in workspace repair.
**Agents/config:** Allow `agents.list[].experimental.localModelLean` so lean local-model mode can be enabled for one configured agent instead of globally.
**Providers/xAI:** Add device-code OAuth login so remote and headless setups can authorize xAI without a localhost browser callback.
**Providers/OpenRouter:** Honor provider-level `params.provider` routing policy for OpenRouter requests, with model and agent params overriding the defaults.
🐛 Bug Fixes
**CLI:** Include stale-running task maintenance decisions in `openclaw tasks maintenance --json` so retained and reconcile candidates explain backing-session, cron, CLI, and wedged-subagent state.
Codex app-server: keep system-prompt reports working when bootstrap hooks provide workspace files with only a path and content, so hook-supplied SOUL/IDENTITY/TOOLS/USER context still reports injected characters correctly.
**Providers/MiniMax music:** Stop advertising `durationSeconds` control and remove prompt-injected duration hints, so `music_generate` reports MiniMax duration as an unsupported override instead of suggesting MiniMax can enforce track length. Fixes #84508.
**Doctor:** Warn when sandbox tool policy hides configured MCP server tools before provider requests.
**WhatsApp:** Update Baileys to `7.0.0-rc12`.
**UI:** Suppress per-locale `rolldown-plugin-dts:fake-js` CommonJS dts warnings emitted while bundling the intentionally-inlined `zod/v4/locales/*.d.cts` files, so `pnpm build` output stays readable after the 0.25.1 plugin bump.
**CLI:** Route lazy plugin-registration logs to stderr for JSON-mode `openclaw nodes` commands so stdout stays parseable.
**Approvals:** Route manual `/approve` decisions through the trusted approval runtime so active exec and plugin approvals no longer look unknown or expired.
**Mac app:** Update the About settings copyright year to 2026.
**Dependencies:** Update `@openclaw/fs-safe` to `0.2.7` so OpenClaw's default Python-helper-off policy keeps best-effort Node write fallbacks for private stores, secret writes, run logs, and media attachments on Linux/macOS.
**Exec approvals:** Remove the old `cat SKILL.md && printf ... && ` allowlist compatibility path so skill files must be loaded with the read tool and only the real skill executable is auto-allowed.
**Discord:** Let voice sessions follow configured Discord users into voice channels, with allowed-channel checks, multi-user handoff, bounded reconciliation, and DAVE recovery preservation.
**Discord/voice:** Include bounded `IDENTITY.md`, `USER.md`, and `SOUL.md` profile context in realtime voice session instructions by default, with `voice.realtime.bootstrapContextFiles: []` available to disable it.
**Dependencies:** Bump the bundled Codex harness to `@openai/codex` `0.132.0` and refresh the app-server model-list docs for the new catalog.
**CLI:** Add the bundled Policy plugin for policy-backed channel conformance checks, doctor lint findings, and opt-in workspace repair.
**Agents/config:** Allow `agents.list[].experimental.localModelLean` so lean local-model mode can be enabled for one configured agent instead of globally.
**Providers/xAI:** Add device-code OAuth login so remote and headless setups can authorize xAI without a localhost browser callback.
**Providers/OpenRouter:** Honor provider-level `params.provider` routing policy for OpenRouter requests, with model and agent params overriding the defaults.
🐛 Bug Fixes
**CLI:** Include stale-running task maintenance decisions in `openclaw tasks maintenance --json` so retained and reconcile candidates explain backing-session, cron, CLI, and wedged-subagent state.
Codex app-server: keep system-prompt reports working when bootstrap hooks provide workspace files with only a path and content, so hook-supplied SOUL/IDENTITY/TOOLS/USER context still reports injected characters correctly.
**Providers/MiniMax music:** Stop advertising `durationSeconds` control and remove prompt-injected duration hints, so `music_generate` reports MiniMax duration as an unsupported override instead of suggesting MiniMax can enforce track length. Fixes #84508.
**Doctor:** Warn when sandbox tool policy hides configured MCP server tools before provider requests.
**WhatsApp:** Update Baileys to `7.0.0-rc12`.
**UI:** Suppress per-locale `rolldown-plugin-dts:fake-js` CommonJS dts warnings emitted while bundling the intentionally-inlined `zod/v4/locales/*.d.cts` files, so `pnpm build` output stays readable after the 0.25.1 plugin bump.
**CLI:** Route lazy plugin-registration logs to stderr for JSON-mode `openclaw nodes` commands so stdout stays parseable.
**Approvals:** Route manual `/approve` decisions through the trusted approval runtime so active exec and plugin approvals no longer look unknown or expired.
**Mac app:** Update the About settings copyright year to 2026.
**Dependencies:** Update `@openclaw/fs-safe` to `0.2.7` so OpenClaw's default Python-helper-off policy keeps best-effort Node write fallbacks for private stores, secret writes, run logs, and media attachments on Linux/macOS.
**Discord:** Let voice sessions follow configured Discord users into voice channels, with allowed-channel checks, multi-user handoff, bounded reconciliation, and DAVE recovery preservation.
**Discord/voice:** Include bounded `IDENTITY.md`, `USER.md`, and `SOUL.md` profile context in realtime voice session instructions by default, with `voice.realtime.bootstrapContextFiles: []` available to disable it.
**Dependencies:** Bump the bundled Codex harness to `@openai/codex` `0.132.0` and refresh the app-server model-list docs for the new catalog.
**CLI:** Add the bundled Policy plugin for policy-backed channel conformance checks, doctor lint findings, and opt-in workspace repair.
**Agents/config:** Allow `agents.list[].experimental.localModelLean` so lean local-model mode can be enabled for one configured agent instead of globally.
**Providers/xAI:** Add device-code OAuth login so remote and headless setups can authorize xAI without a localhost browser callback.
**Providers/OpenRouter:** Honor provider-level `params.provider` routing policy for OpenRouter requests, with model and agent params overriding the defaults.
🐛 Bug Fixes
**WhatsApp:** Update Baileys to `7.0.0-rc12`.
**Dependencies:** Update `@openclaw/fs-safe` to `0.2.7` so OpenClaw's default Python-helper-off policy keeps best-effort Node write fallbacks for private stores, secret writes, run logs, and media attachments on Linux/macOS.
**Browser:** Honor the configured image sanitization limit for screenshots and labeled snapshots so browser-captured images follow the same resize policy as other image results.
**Doctor:** Remove unrecognized `models.providers.*.models[*].compat.thinkingFormat` values during `doctor --fix` so stale provider model config can validate after upgrade. Fixes #77803.
**Status:** Show the configured default, session-selected model, reason, clear hint, and docs link when a session remains pinned to a model that differs from `agents.defaults.model.primary`.
**Mac app:** Keep local packaging signed with a stable app identity for permission testing and fix Control UI production builds under current Vite/Highlight.js exports.
**macOS app:** Update the embedded Peekaboo bridge to 3.2.1 so OpenClaw-hosted UI automation works with current Peekaboo CLI capture flows.
**Cron:** Deliver preferred final assistant output for successful scheduled runs when trailing plain tool warnings remain in diagnostics instead of marking the run failed.
Fix(mattermost): fail closed on missing channel type [AI].
**Agents:** Clarify that fixes should default to clean bounded refactors, lean internals, and explicit plugin SDK/API deprecation paths.
**Dependencies:** Update `@openclaw/proxyline` to 0.3.3.
**Dependencies:** Update Pi packages to 0.75.1 and raise the minimum supported Node.js 22 line to 22.19.
**Docker/Podman:** Add `OPENCLAW_IMAGE_APT_PACKAGES` as the runtime-neutral image build arg for extra apt packages while keeping `OPENCLAW_DOCKER_APT_PACKAGES` as a legacy fallback.
**Gateway/ACPX:** Attribute startup probe, config, runtime, and resource-count costs in restart traces without changing readiness behavior.
**Gateway:** Overlap startup logging and plugin-service startup with channel sidecars to reduce restart ready latency while preserving `/readyz` sidecar gating.
Plugins/admin-http-rpc: You can now trusted admin HTTP RPC clients to start and wait for web QR login flows.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Mac app:** Refine Voice & Talk recognition-language and wake-phrase settings so they use the same compact card rows as the rest of Settings.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
...and 33 more
🐛 Bug Fixes
**CLI:** Reject explicit port numbers above 65535 before they reach Gateway or Node bind paths. Fixes #83900.
Codex app-server: preserve plugin tool auth profiles when Codex owns model transport so OpenClaw dynamic tools can **Fixed:** Their provider credentials.
**Memory/search:** Scan the JS-side fallback vector path (used when the sqlite-vec index is unavailable or has a mismatched dimension) in bounded rowid batches and yield to the event loop between batches so large chunk tables can no longer pin the Node.js main thread for multi-second windows. Also keeps the SQL prepared statement rooted in a local so node:sqlite cannot finalize it mid-scan under heap pressure. Fixes #81172.
**Memory Wiki:** Preserve fs-safe diagnostics when bridge source page writes fail for non-symlink filesystem safety reasons, so directory collisions are reported with the underlying error code.
**Telegram:** Keep forum topics from blocking sibling topic traffic by routing inbound serialization, media/text buffers, and account API queues on topic-aware lanes.
**Telegram:** Keep queued forum-topic follow-up messages from inheriting superseded source abort signals, so later same-topic user turns can still run and reply after an active turn is replaced.
**CLI:** Bypass npm freshness filters consistently during managed package and plugin installs so freshly published release plugins remain installable.
**CLI:** Guide root-owned npm install EACCES recovery by stopping the managed Gateway before manual package replacement, then reinstalling and restarting the service. Fixes #83747.
**Twitch:** Register refreshing chat tokens with Twurple's chat intent so automatic token refresh keeps chat access available.
**Agents/subagents:** Keep collect-mode announce queues batching unresolved-origin items with compatible same-route messages and resume collection after a true cross-channel drain when a later compatible batch remains. Fixes #83577.
**Agents:** Clarify that fixes should default to clean bounded refactors, lean internals, and explicit plugin SDK/API deprecation paths.
**Dependencies:** Update `@openclaw/proxyline` to 0.3.3.
**Dependencies:** Update Pi packages to 0.75.1 and raise the minimum supported Node.js 22 line to 22.19.
**Docker/Podman:** Add `OPENCLAW_IMAGE_APT_PACKAGES` as the runtime-neutral image build arg for extra apt packages while keeping `OPENCLAW_DOCKER_APT_PACKAGES` as a legacy fallback.
**Gateway/ACPX:** Attribute startup probe, config, runtime, and resource-count costs in restart traces without changing readiness behavior.
**Gateway:** Overlap startup logging and plugin-service startup with channel sidecars to reduce restart ready latency while preserving `/readyz` sidecar gating.
Plugins/admin-http-rpc: You can now trusted admin HTTP RPC clients to start and wait for web QR login flows.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Mac app:** Refine Voice & Talk recognition-language and wake-phrase settings so they use the same compact card rows as the rest of Settings.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
...and 33 more
🐛 Bug Fixes
**CLI:** Reject explicit port numbers above 65535 before they reach Gateway or Node bind paths. Fixes #83900.
Codex app-server: preserve plugin tool auth profiles when Codex owns model transport so OpenClaw dynamic tools can **Fixed:** Their provider credentials.
**Memory/search:** Scan the JS-side fallback vector path (used when the sqlite-vec index is unavailable or has a mismatched dimension) in bounded rowid batches and yield to the event loop between batches so large chunk tables can no longer pin the Node.js main thread for multi-second windows. Also keeps the SQL prepared statement rooted in a local so node:sqlite cannot finalize it mid-scan under heap pressure. Fixes #81172.
**Memory Wiki:** Preserve fs-safe diagnostics when bridge source page writes fail for non-symlink filesystem safety reasons, so directory collisions are reported with the underlying error code.
**Telegram:** Keep forum topics from blocking sibling topic traffic by routing inbound serialization, media/text buffers, and account API queues on topic-aware lanes.
**Telegram:** Keep queued forum-topic follow-up messages from inheriting superseded source abort signals, so later same-topic user turns can still run and reply after an active turn is replaced.
**CLI:** Bypass npm freshness filters consistently during managed package and plugin installs so freshly published release plugins remain installable.
**CLI:** Guide root-owned npm install EACCES recovery by stopping the managed Gateway before manual package replacement, then reinstalling and restarting the service. Fixes #83747.
**Twitch:** Register refreshing chat tokens with Twurple's chat intent so automatic token refresh keeps chat access available.
**Agents/subagents:** Keep collect-mode announce queues batching unresolved-origin items with compatible same-route messages and resume collection after a true cross-channel drain when a later compatible batch remains. Fixes #83577.
**Agents:** Clarify that fixes should default to clean bounded refactors, lean internals, and explicit plugin SDK/API deprecation paths.
**Dependencies:** Update `@openclaw/proxyline` to 0.3.3.
**Dependencies:** Update Pi packages to 0.75.1 and raise the minimum supported Node.js 22 line to 22.19.
**Docker/Podman:** Add `OPENCLAW_IMAGE_APT_PACKAGES` as the runtime-neutral image build arg for extra apt packages while keeping `OPENCLAW_DOCKER_APT_PACKAGES` as a legacy fallback.
**Gateway/ACPX:** Attribute startup probe, config, runtime, and resource-count costs in restart traces without changing readiness behavior.
**Gateway:** Overlap startup logging and plugin-service startup with channel sidecars to reduce restart ready latency while preserving `/readyz` sidecar gating.
Plugins/admin-http-rpc: You can now trusted admin HTTP RPC clients to start and wait for web QR login flows.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Mac app:** Refine Voice & Talk recognition-language and wake-phrase settings so they use the same compact card rows as the rest of Settings.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
...and 33 more
🐛 Bug Fixes
**CLI:** Reject explicit port numbers above 65535 before they reach Gateway or Node bind paths. Fixes #83900.
Codex app-server: preserve plugin tool auth profiles when Codex owns model transport so OpenClaw dynamic tools can **Fixed:** Their provider credentials.
**Memory/search:** Scan the JS-side fallback vector path (used when the sqlite-vec index is unavailable or has a mismatched dimension) in bounded rowid batches and yield to the event loop between batches so large chunk tables can no longer pin the Node.js main thread for multi-second windows. Also keeps the SQL prepared statement rooted in a local so node:sqlite cannot finalize it mid-scan under heap pressure. Fixes #81172.
**Memory Wiki:** Preserve fs-safe diagnostics when bridge source page writes fail for non-symlink filesystem safety reasons, so directory collisions are reported with the underlying error code.
**Telegram:** Keep forum topics from blocking sibling topic traffic by routing inbound serialization, media/text buffers, and account API queues on topic-aware lanes.
**Telegram:** Keep queued forum-topic follow-up messages from inheriting superseded source abort signals, so later same-topic user turns can still run and reply after an active turn is replaced.
**CLI:** Bypass npm freshness filters consistently during managed package and plugin installs so freshly published release plugins remain installable.
**CLI:** Guide root-owned npm install EACCES recovery by stopping the managed Gateway before manual package replacement, then reinstalling and restarting the service. Fixes #83747.
**Twitch:** Register refreshing chat tokens with Twurple's chat intent so automatic token refresh keeps chat access available.
**Agents/subagents:** Keep collect-mode announce queues batching unresolved-origin items with compatible same-route messages and resume collection after a true cross-channel drain when a later compatible batch remains. Fixes #83577.
**Agents:** Clarify that fixes should default to clean bounded refactors, lean internals, and explicit plugin SDK/API deprecation paths.
**Dependencies:** Update `@openclaw/proxyline` to 0.3.3.
**Dependencies:** Update Pi packages to 0.75.1 and raise the minimum supported Node.js 22 line to 22.19.
**Docker/Podman:** Add `OPENCLAW_IMAGE_APT_PACKAGES` as the runtime-neutral image build arg for extra apt packages while keeping `OPENCLAW_DOCKER_APT_PACKAGES` as a legacy fallback.
**Gateway/ACPX:** Attribute startup probe, config, runtime, and resource-count costs in restart traces without changing readiness behavior.
**Gateway:** Overlap startup logging and plugin-service startup with channel sidecars to reduce restart ready latency while preserving `/readyz` sidecar gating.
Plugins/admin-http-rpc: You can now trusted admin HTTP RPC clients to start and wait for web QR login flows.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
**Skills:** Add a meme-maker skill for curated template search, local SVG/PNG rendering, Imgflip hosted rendering, and Know Your Meme provenance links.
...and 29 more
🐛 Bug Fixes
**Memory/search:** Scan the JS-side fallback vector path (used when the sqlite-vec index is unavailable or has a mismatched dimension) in bounded rowid batches and yield to the event loop between batches so large chunk tables can no longer pin the Node.js main thread for multi-second windows. Also keeps the SQL prepared statement rooted in a local so node:sqlite cannot finalize it mid-scan under heap pressure. Fixes #81172.
**CLI:** Bypass npm freshness filters consistently during managed package and plugin installs so freshly published release plugins remain installable.
**Agents/subagents:** Keep collect-mode announce queues batching unresolved-origin items with compatible same-route messages and resume collection after a true cross-channel drain when a later compatible batch remains. Fixes #83577.
**Providers/Anthropic:** Preserve native image input for current Claude model rows when stale local catalog data marks them text-only.
**UI:** Render live tool progress from session-scoped `session.tool` Gateway events so externally started runs show their tool cards in the active session.
**Outbound:** Resolve send-capable channel plugins from the active runtime registry when the pinned startup registry only has setup metadata.
**Browser:** Enforce current-tab URL allowlist checks for `/act` evaluate/batch actions and `/highlight` routes while leaving tab-management actions unblocked.
**CI:** Require real-behavior-proof verdict markers to come from the ClawSweeper GitHub App before accepting exact-head proof.
**Models:** Show the effective OpenAI/Codex auth profile in `/models` provider headers instead of falling back to the OpenAI env-key label.
**Browser:** Keep a profile `cdpPort` when its `cdpUrl` omits a port, while still letting explicitly written URL ports win.
**Agents:** Clarify that fixes should default to clean bounded refactors, lean internals, and explicit plugin SDK/API deprecation paths.
**Dependencies:** Update `@openclaw/proxyline` to 0.3.3.
**Dependencies:** Update Pi packages to 0.75.1 and raise the minimum supported Node.js 22 line to 22.19.
**Docker/Podman:** Add `OPENCLAW_IMAGE_APT_PACKAGES` as the runtime-neutral image build arg for extra apt packages while keeping `OPENCLAW_DOCKER_APT_PACKAGES` as a legacy fallback.
**Gateway/ACPX:** Attribute startup probe, config, runtime, and resource-count costs in restart traces without changing readiness behavior.
**Gateway:** Overlap startup logging and plugin-service startup with channel sidecars to reduce restart ready latency while preserving `/readyz` sidecar gating.
Plugins/admin-http-rpc: You can now trusted admin HTTP RPC clients to start and wait for web QR login flows.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
**Skills:** Add a meme-maker skill for curated template search, local SVG/PNG rendering, Imgflip hosted rendering, and Know Your Meme provenance links.
...and 21 more
🐛 Bug Fixes
**Discord/OpenAI:** Keep realtime Discord voice sessions hearing follow-up turns with OpenAI realtime and prebuffer assistant playback to avoid choppy starts.
**Media:** Prevent image metadata probing from invoking external decoder delegates on unrecognized image bytes, and stop fallback chaining after real processing errors.
**Media:** Install Sharp with the root package and fall back to sips, Windows native imaging, ImageMagick, GraphicsMagick, or ffmpeg for image resizing/conversion when Sharp is unavailable. Fixes #83401.
**Telegram:** Deliver generated media completions back into forum topics by preserving topic IDs across requester-agent handoff.
**Gateway:** Defer update-check startup until after readiness so package update checks no longer block sidecar-ready startup, while preserving update broadcasts and shutdown cleanup.
**Telegram:** Keep `/btw` and read-only status commands from aborting active runs, and avoid retaining raw update payloads in timed-out spool tombstones. Refs #83272.
**Agents/video:** Hide `video_generate` reference-audio parameters unless a registered video provider supports audio inputs.
**Plugins/xAI:** Echo PKCE challenge fields during OAuth authorization-code token exchange for xAI token-endpoint compatibility.
Codex app-server: hydrate current inbound image attachments before queued runs so Responses-backed agents receive Discord and other channel images as native vision input. Fixes #83466.
Codex app-server: keep native code mode available without forcing code-mode-only so OpenClaw dynamic tool turns complete through the app-server tool bridge. Fixes #83109.
**Dependencies:** Update `@openclaw/proxyline` to 0.3.3.
**Dependencies:** Update Pi packages to 0.75.1 and raise the minimum supported Node.js 22 line to 22.19.
**Docker/Podman:** Add `OPENCLAW_IMAGE_APT_PACKAGES` as the runtime-neutral image build arg for extra apt packages while keeping `OPENCLAW_DOCKER_APT_PACKAGES` as a legacy fallback.
**Gateway/ACPX:** Attribute startup probe, config, runtime, and resource-count costs in restart traces without changing readiness behavior.
**Gateway:** Overlap startup logging and plugin-service startup with channel sidecars to reduce restart ready latency while preserving `/readyz` sidecar gating.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
**Skills:** Add a meme-maker skill for curated template search, local SVG/PNG rendering, Imgflip hosted rendering, and Know Your Meme provenance links.
**Browser:** Surface pending and recently handled modal dialogs in snapshots, return `blockedByDialog` when an action opens a modal, and allow `browser dialog --dialog-id` to answer pending dialogs.
**Agents/tools:** Shorten built-in tool descriptions and schema hints across media, messaging, sessions, cron, Gateway, web, image/PDF, TTS, nodes, and plan tools while preserving routing guardrails.
...and 17 more
🐛 Bug Fixes
**Agents/code mode:** Preserve agent, session, run, and channel context in `before_tool_call` hooks for top-level `exec`/`wait` dispatches. Fixes #83387.
**Replies:** Keep final payload delivery after live preview updates so channels can finalize or send the completed answer instead of losing preview-only drafts.
**Providers/Xiaomi:** Replay MiMo Anthropic-compatible `reasoning_content` as provider-required thinking blocks even when OpenClaw thinking is disabled, fixing follow-up tool turns for `mimo-v2-flash`. Fixes #83407.
**Agents/exec approvals:** Forward approval-runtime credentials on agent-owned Gateway approval calls so approved async commands complete through the existing runtime path instead of stalling on unauthenticated follow-up calls. Thanks @IWhatsskill, @Patrick-Erichsen, and @jesse-merhi.
**Gateway/skills:** Preflight remote macOS skill-bin refreshes with a WebSocket connectivity check so stale node sessions skip quickly instead of logging slow `system.which` timeout warnings.
**CLI:** Keep broken discovered plugins that are not referenced by active config from failing `openclaw config validate`, while preserving fatal errors for explicitly configured plugin entries.
**GitHub Copilot:** Drop unsafe native Responses reasoning replay items with non-replayable IDs before dispatch, preventing affected Copilot sessions from failing with `invalid_request_body`. Fixes #83220.
**Agents/Codex:** Fail closed when an explicitly requested Codex harness is not registered instead of silently trying configured model fallbacks. Fixes #83349.
QA-Lab: make runtime tool coverage fail on missing required tool exercise instead of treating pass/pass parity envelope drift as missing coverage.
**Core/plugins:** Harden clawpatch-reported edge cases across gateway auth cleanup, Claude session id paths, plugin activation policy, apply-patch hunk handling, diagnostic redaction, and plugin metadata validation.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
**Skills:** Add a meme-maker skill for curated template search, local SVG/PNG rendering, Imgflip hosted rendering, and Know Your Meme provenance links.
**Browser:** Surface pending and recently handled modal dialogs in snapshots, return `blockedByDialog` when an action opens a modal, and allow `browser dialog --dialog-id` to answer pending dialogs.
**Agents/tools:** Shorten built-in tool descriptions and schema hints across media, messaging, sessions, cron, Gateway, web, image/PDF, TTS, nodes, and plan tools while preserving routing guardrails.
**CLI:** Add `defineToolPlugin` plus `openclaw plugins build`, `validate`, and `init` for typed simple tool plugins with generated manifest metadata, optional tool declarations, and context factories.
**Agents/skills:** Tighten bundled skill prompts and metadata, quote skill descriptions, refresh current CLI/API guidance, and update embedded sherpa-onnx runtime downloads.
**Skills:** Update the Obsidian skill to target the official `obsidian` CLI and require its registered binary instead of the third-party `obsidian-cli`.
**Skills:** Add a Python debugging skill for pdb, breakpoint(), post-mortem inspection, and debugpy remote attach.
...and 9 more
🐛 Bug Fixes
**Mac app:** Keep app-level menu commands and Dashboard failure states reachable when the remote Gateway is disconnected, and keep the Settings sidebar toggle in the leading titlebar area.
**UI:** Hide internal runtime-context and other `display: false` transcript messages from Chat history and live message events. Fixes #83216.
**Mac app:** Align the Sessions settings pane with the standard Settings page gutter and row spacing.
**OpenAI/Codex:** Stop rejecting available `openai-codex` GPT-5.1, GPT-5.2, and GPT-5.3 model refs during config validation, while keeping removed Spark aliases suppressed. Fixes #83303.
Codex app-server: preserve streamed native command output in mirrored transcripts and trajectory exports when final snapshots omit aggregated output.
Codex app-server: fail closed when chat or sender policy denies tools, disabling native code, app, environment, and user MCP surfaces for restricted turns.
Codex app-server: keep recent context-engine messages when oversized projected history is truncated, so short follow-ups in long channel sessions do not fall back to stale earlier turns.
**Feishu:** Return bound subagent delivery origins from session thread setup so Feishu subagent completions route back to the same DM or topic.
**CLI:** Tailor post-update Gateway recovery hints by platform, showing systemd, LaunchAgent, Scheduled Task, or generic service-manager guidance instead of macOS-only recovery text.
**Plugins:** Apply a default 15-second timeout to legacy `before_agent_start` hooks so hung plugin handlers no longer block agent startup. Fixes #48534.
**Mac app:** Redesign Settings pages with consistent card layouts, cached navigation, cleaner permissions/voice/skills/cron/exec/debug panes, and steadier spacing around the native sidebar.
**Skills:** Rename the repo-local Codex closeout review skill and helper to `autoreview` while preserving the Codex-first fallback behavior.
**Skills:** Add a meme-maker skill for curated template search, local SVG/PNG rendering, Imgflip hosted rendering, and Know Your Meme provenance links.
**Agents/tools:** Shorten built-in tool descriptions and schema hints across media, messaging, sessions, cron, Gateway, web, image/PDF, TTS, nodes, and plan tools while preserving routing guardrails.
**CLI:** Add `defineToolPlugin` plus `openclaw plugins build`, `validate`, and `init` for typed simple tool plugins with generated manifest metadata, optional tool declarations, and context factories.
**Agents/skills:** Tighten bundled skill prompts and metadata, quote skill descriptions, refresh current CLI/API guidance, and update embedded sherpa-onnx runtime downloads.
**Skills:** Update the Obsidian skill to target the official `obsidian` CLI and require its registered binary instead of the third-party `obsidian-cli`.
**Skills:** Add a Python debugging skill for pdb, breakpoint(), post-mortem inspection, and debugpy remote attach.
**Plugins/messages:** Add presentation capability limits for channel renderers, adapt rich message controls before native rendering, and mark legacy `interactive`/Slack directive producer APIs as deprecated.
...and 8 more
🐛 Bug Fixes
**CLI:** Tailor post-update Gateway recovery hints by platform, showing systemd, LaunchAgent, Scheduled Task, or generic service-manager guidance instead of macOS-only recovery text.
**Plugins:** Apply a default 15-second timeout to legacy `before_agent_start` hooks so hung plugin handlers no longer block agent startup. Fixes #48534.
**Feishu:** Refresh inbound session delivery context for DM, group, and broadcast turns so later replies do not inherit stale WebChat routing. Fixes #78274.
QA-Lab/qa-channel: attach redacted agent tool-start traces to outbound `QaBusMessage` records so scenarios can assert actual tool use instead of relying only on reply text. Fixes #67637.
QA-Lab: fail live runtime parity reports when assistant-message usage is missing, preventing `0 vs 0` live token rows from being reported as passing proof. Fixes #80411.
QA-Lab: fail Codex-backed OpenAI live runtime-pair runs before launching isolated workers when no portable Codex auth is available, while staging API-key fallbacks and configured Codex keys for isolated QA agents. Fixes #80412.
QA-Lab: refresh parity gates, mock frontier fixtures, model scenarios, and workflow artifact lanes to compare GPT-5.5 against Claude Opus 4.7. Fixes #74262.
QA-Lab: make mock parity dispatch provider-aware for source discovery and subagent scenarios so OpenAI and Anthropic lanes no longer share identical canned plans. Fixes #64879.
QA-Lab: **Fixed:** Returning Control UI bearer tokens from unauthenticated bootstrap payloads and bind Docker harness ports to loopback-only host addresses.
**Mac app:** Avoid a SwiftUI metadata crash when rendering the Cron Jobs settings pane.
**🔒 Security/audit:** Add `security.audit.suppressions` for intentionally accepted audit findings, keeping suppressed matches out of the active summary while preserving them in JSON output with an active suppression notice.
**Agents/subagents:** Label delegated task and subagent completion handoffs as ready for parent review, and tell requester agents to review/verify results before calling them done.
**Providers/media:** Add fal and OpenRouter music-generation providers for the shared `music_generate` tool, including fal MiniMax/ACE/Stable Audio endpoints and OpenRouter Lyria audio output.
**Maintainer tooling:** Warn before running JS package commands on raw Crabbox AWS boxes, pointing maintainers to Actions hydration or Blacksmith Testbox for CI-like proof.
**UI:** Show provider quota usage in the Overview card and Chat header, and recover stale Chat in-progress state after missed terminal events.
Mac app remote setup can now be preconfigured from `openclaw-mac configure-remote`, skips onboarding when config is already complete, supports direct LAN/Tailnet gateway URLs, allows private same-origin Control UI loads, and owns the SSH tunnel process when SSH is selected.
**Providers/xAI:** Add xAI Grok OAuth login for SuperGrok subscribers, letting `xai/*` models and xAI media/tool providers authenticate without `XAI_API_KEY`.
**CLI:** Add `openclaw cron run --wait` with timeout and poll interval controls, plus exact `cron.runs --run-id` filtering so automation can block on one queued manual run.
**Maintainer tooling:** Route Crabbox skill defaults through the repo brokered AWS config, leaving Blacksmith Testbox as an explicit opt-in instead of the broad-proof default.
**CLI:** Localize the setup wizard and bundled channel setup flows for English, Simplified Chinese, and Traditional Chinese.
...and 10 more
🐛 Bug Fixes
**Agents/subagents:** Route group/channel subagent completions through message-tool-only handoffs when required and keep active-requester wake failures from dropping completion delivery. Fixes #82803. Thanks @galiniliev, @yozakura-ava, and @moeedahmed.
Memory-core: scan persisted memory source sessions on startup, comparing on-disk transcripts against the index and marking only missing/newer/resized files dirty for incremental sync. Fixes #82341.
**Telegram:** Keep the top-level default account in the account list when named accounts or bindings are added alongside top-level credentials, preserving default polling while still letting named-only configs resolve to a single account. Fixes #82794.
**WhatsApp:** Honor forced document delivery for outbound image, GIF, and video media so `forceDocument`/`asDocument` sends preserve original media bytes instead of using compressed media payloads.
**Gateway/usage:** Refresh large session usage summaries in the background and reuse durable transcript metadata so `sessions.usage` no longer blocks Gateway requests on full transcript rescans. Fixes #82773.
**UI:** Restore the submitted draft when chat is busy instead of clearing it or queueing another run. Fixes #45326.
**Browser plugin:** Redact attach-details from Chrome MCP diagnostics and keep raw Chrome launch error output around long enough to surface in user reports without leaking sensitive paths.
**System prompts:** Clarify MEMORY guidance over generic TTS hints in the embedded speech-core/system-prompt scaffolding so agents prefer memory-store usage over speech defaults. Fixes #81930.
**🔒 Agents/auth:** Include the checked credential source in missing API key errors, so users can see which env var, profile, or config path to fix. Fixes #82785.
**Providers/GitHub Copilot:** Hash Responses replay item ids with sha256 instead of a weak 32-bit hash and build same-provider Copilot tool-call ids distinctly, so concurrent tool-call replays no longer collide and reject follow-up turns.
**Providers/xAI:** Add xAI Grok OAuth login for SuperGrok subscribers, letting `xai/*` models and xAI media/tool providers authenticate without `XAI_API_KEY`.
**CLI:** Add `openclaw cron run --wait` with timeout and poll interval controls, plus exact `cron.runs --run-id` filtering so automation can block on one queued manual run.
**Maintainer tooling:** Route Crabbox skill defaults through the repo brokered AWS config, leaving Blacksmith Testbox as an explicit opt-in instead of the broad-proof default.
**CLI:** Localize the setup wizard and bundled channel setup flows for English, Simplified Chinese, and Traditional Chinese.
**Agents/skills:** Cache hydrated `resolvedSkills` across warm gateway turns while keying reuse by the redacted effective config, reducing redundant skill snapshot rebuilds without crossing config-gated skill boundaries.
**Telegram/group chat:** Add opt-in `messages.groupChat.ambientTurns: "room_event"` handling so always-on ambient chatter can run as quiet room context and speak visibly only via the message tool.
**Codex/context engines:** Bind thread-bootstrap projection epochs to Codex app-server threads, carry redacted tool-result context into fresh threads, and rotate backend threads when projection state changes.
**Gateway:** Add opt-in restart trace logs for restart signal, active-work drain, close, next-start, ready, and memory spans.
**Gateway/performance:** Split startup benchmark HTTP-listen timing from full gateway-ready timing and add post-bind plugin and sidecar diagnostics to restart-readiness traces.
🐛 Bug Fixes
**Gateway/diagnostics:** Redact credential-bearing gateway target URLs and client diagnostics while preserving raw connection URLs for programmatic use, so connect-failure logs no longer surface embedded tokens.
**Telegram:** Normalize announce group targets via a new `resolveSessionTarget` channel hook so scheduled announcements resolve consistently against the same Telegram session conversation registry as inbound turns. Fixes #81229.
**Discord:** Bind delayed gateway `identify` retries to the originating socket generation so retries triggered after a reconnect do not identify against a fresh socket. Fixes #82225.
**UI:** Refresh cached runtime handles when agent config changes so ACP sessions stop using stale runtimes after `agents.defaults` edits. Fixes #82237.
**Gateway/sessions:** Scope session data lookups by agent id so multi-agent gateway state cannot cross-leak session records across configured agents.
**Agents/media:** Accept generated media attachments on internal completion events and report delivery-loss failures as errors, so completed background music/video tasks do not disappear after provider success.
**Matrix/approvals:** Release in-flight reaction bindings when the channel approval handler stops mid-delivery, preventing stale approval targets after restart. Fixes #82485.
Matrix/E2EE: **Fixed:** Requesting MSC4222 `state_after` sync responses so homeservers with incomplete state-after data do not leave fresh encrypted rooms without outbound room encryptors. Fixes #82515.
**UI:** Update the displayed model in real time when an auto-fallback resolution swaps in a different model mid-turn, so the status line reflects the actual model handling the run. Fixes #82296.
**Gateway/sessions:** Estimate context usage from local/OpenAI-compatible transcripts when provider usage telemetry is missing, so status no longer shows empty usage for real local-model sessions. Fixes #73990.
**Providers/xAI:** Add xAI Grok OAuth login for SuperGrok subscribers, letting `xai/*` models and xAI media/tool providers authenticate without `XAI_API_KEY`.
**CLI:** Add `openclaw cron run --wait` with timeout and poll interval controls, plus exact `cron.runs --run-id` filtering so automation can block on one queued manual run.
**Maintainer tooling:** Route Crabbox skill defaults through the repo brokered AWS config, leaving Blacksmith Testbox as an explicit opt-in instead of the broad-proof default.
**CLI:** Localize the setup wizard and bundled channel setup flows for English, Simplified Chinese, and Traditional Chinese.
**Agents/skills:** Cache hydrated `resolvedSkills` across warm gateway turns while keying reuse by the redacted effective config, reducing redundant skill snapshot rebuilds without crossing config-gated skill boundaries.
**Telegram/group chat:** Add opt-in `messages.groupChat.ambientTurns: "room_event"` handling so always-on ambient chatter can run as quiet room context and speak visibly only via the message tool.
**Codex/context engines:** Bind thread-bootstrap projection epochs to Codex app-server threads, carry redacted tool-result context into fresh threads, and rotate backend threads when projection state changes.
🐛 Bug Fixes
**Telegram:** Persist polling updates through restart replay so queued same-topic messages resume in order instead of losing context after a gateway restart.
**Gateway/Gmail:** Abort in-flight Gmail watcher startup and hot-reload restarts before shutdown so reloads cannot spawn `gog serve` after the Gateway is closing.
**MCP plugin tools:** Forward host MCP `tools/call` `AbortSignal` through `createPluginToolsMcpHandlers().callTool` into plugin `tool.execute`, so host cancellation actually cancels in-flight plugin tool calls instead of letting them run to completion. Fixes #82424.
**Media:** Ignore image MIME and filename hints when bytes sniff as generic containers, so zip/octet-stream payloads mislabeled as images do not become local image media or keep image file extensions when staged.
**Update/doctor:** Avoid materializing `groupAllowFrom` for channel schemas that reject it, so package-swap doctor repairs do not fail on externalized Slack configs.
**Gateway/media:** Prevent image filenames from overriding generic non-image byte sniffing, so zip/octet-stream payloads mislabeled as images are offloaded or rejected before they become inline image attachments.
**UI:** Downgrade stale optional provider installs to warnings so Gateway and doctor repair paths keep running after startup provider selection. Refs #82313.
**Telegram/Gateway:** Route targeted Telegram `/stop@bot` messages onto the control lane without cached bot metadata and match gateway stop requests across raw/canonical session aliases.
**MS Teams/media:** Sniff inline `data:image/*` attachment bytes before staging them, skipping payloads that are not actually images.
**Update:** Let package-swap `doctor --fix` persist core config repairs while plugin schemas are still converging, preventing update failures on externalized channel configs.
**Maintainer tooling:** Route Crabbox skill defaults through the repo brokered AWS config, leaving Blacksmith Testbox as an explicit opt-in instead of the broad-proof default.
**CLI:** Localize the setup wizard and bundled channel setup flows for English, Simplified Chinese, and Traditional Chinese.
**Agents/skills:** Cache hydrated `resolvedSkills` across warm gateway turns while keying reuse by the redacted effective config, reducing redundant skill snapshot rebuilds without crossing config-gated skill boundaries.
**Telegram/group chat:** Add opt-in `messages.groupChat.ambientTurns: "room_event"` handling so always-on ambient chatter can run as quiet room context and speak visibly only via the message tool.
🐛 Bug Fixes
**Release tooling:** Align the published launcher Node floor, `npm start`, package script checks, sharded lint locking, Vitest root project coverage, and plugin-SDK declaration build cache metadata so release/package validation does not silently skip or ship stale surfaces.
**Cron/agents:** Honor configured subagent model fallbacks for isolated scheduled runs and forward that fallback policy into embedded agent timeout failover. Fixes #74985.
Codex app-server/MCP: scope user MCP servers to specific OpenClaw agent ids through an optional `mcp.servers..codex.agents` list and accept `codex.defaultToolsApprovalMode` (`auto`/`prompt`/`approve`) for native Codex approval defaults; OpenClaw strips the `codex` block before handing `mcp_servers` config to Codex.
**Agents/OpenAI Responses:** Clamp `input_tokens - cached_tokens` at zero and reconstruct `totalTokens` from input + output + cached components so Responses-API streams report consistent usage when providers under-report `input_tokens` relative to `cached_tokens`.
**Plugins:** Reject malformed `package.json` `openclaw.extensions` metadata during install, discovery, and post-update payload smoke instead of silently dropping invalid entries.
**Media/files:** Sniff `input_file` bytes before trusting declared MIME headers, rejecting spoofed image or zip payloads before they become agent-visible text.
**Config persistence:** Ignore malformed array/scalar auth profile, cron job state, and session store entries instead of hydrating them into numeric profile ids, crashed cron rows, or invalid session records.
**Providers:** Reject malformed successful Runway, BytePlus, and Ollama embedding responses with provider-owned errors instead of raw parser/type failures, silent bad vectors, or long bogus polling.
**Trajectory export:** Skip and report malformed session/runtime JSONL rows in `manifest.json` instead of letting wrong-shaped session rows crash support bundle export.
**Config/doctor:** Copy fallback-enabled channel `allowFrom` entries into explicit `groupAllowFrom` allowlists during `openclaw doctor --fix`, preserving current group access without adding runtime fallback-transition flags.
**Channels/SDK:** Add normalized command turn facts to channel turn construction and expose command-turn helpers for plugin inbound contexts.
**Agents/config:** Support per-agent bootstrap profile overrides for `contextInjection`, `bootstrapMaxChars`, and `bootstrapTotalMaxChars`, inheriting from `agents.defaults` when omitted. Fixes #69966.
**Dependencies:** Route root ambient Node proxy agents through `@openclaw/proxyline` and drop root `proxy-agent`, `https-proxy-agent`, and `minimatch` dependencies.
**Canvas:** Lazy-load HTTP host, hosted media resolver, CLI implementation, and tool runtime modules so Gateway startup only pays Canvas implementation cost on first use.
Control UI/i18n: add a `pnpm ui:i18n:report` baseline report for hardcoded-copy focus areas and locale fallback metadata.
**Maintainer tooling:** Add a repo-local `codex-review` skill for Codex closeout reviews, including local dirty-work and PR-branch review helpers that rerun until no accepted/actionable findings remain and avoid unsupported inline prompts with `--base`.
**Maintainer tooling:** Fail CI when pull requests add package patch files or pnpm patched dependencies, preserving the upstream-and-bump dependency workflow.
Codex app-server: stream commentary preambles into editable channel progress drafts without promoting them to final answers.
**Codex migration:** Remove the bundled `codex-cli` backend and repair legacy `codex-cli/*` model refs to the Codex app-server route on `openai/*`.
**Gateway/startup:** Add owner-level startup trace attribution for auth, plugin loading, lookup counts, and plugin sidecar services.
...and 16 more
🐛 Bug Fixes
**Models/providers:** Trust the exact configured custom/local provider `baseUrl` origin for guarded model HTTP requests, so loopback, LAN, tailnet, and private DNS endpoints work without broad private-network access while different ports and metadata/link-local pivots remain blocked. Fixes #80732. Thanks @Kaspre and @msitarzewski.
Bind shell script operands after combined options [AI].
**Agents/Azure OpenAI Responses:** Default unset Azure OpenAI API versions to `preview` so `/openai/v1/responses` calls use Azure's current Responses API route.
**UI:** Compact the desktop chat header controls into a single aligned row so the session, model, thinking, and action controls no longer waste vertical space.
**Agents:** Retry empty final turns for generic `anthropic-messages` providers instead of limiting non-visible recovery to Kimi, so custom/proxied Anthropic-compatible routes can recover with a visible answer. Addresses #46080. Thanks @wmgx, @w1tv, and @iFwu.
**Agents/replies:** Strip workflow `` scaffolding from user-visible sanitizer paths so raw tool output does not leak into chat history, transcript mirrors, or channel replies. Fixes #47444.
**Agents/media:** Deliver generated image, music, and video results through structured attachments, keep message-tool-only Codex completions on the message tool, and fail completion handoff when expected media is not actually sent.
**Dependencies:** Route root ambient Node proxy agents through `@openclaw/proxyline` and drop root `proxy-agent`, `https-proxy-agent`, and `minimatch` dependencies.
Control UI/i18n: add a `pnpm ui:i18n:report` baseline report for hardcoded-copy focus areas and locale fallback metadata.
**Maintainer tooling:** Add a repo-local `codex-review` skill for Codex closeout reviews, including local dirty-work and PR-branch review helpers that rerun until no accepted/actionable findings remain and avoid unsupported inline prompts with `--base`.
**Maintainer tooling:** Fail CI when pull requests add package patch files or pnpm patched dependencies, preserving the upstream-and-bump dependency workflow.
Codex app-server: stream commentary preambles into editable channel progress drafts without promoting them to final answers.
**Codex migration:** Remove the bundled `codex-cli` backend and repair legacy `codex-cli/*` model refs to the Codex app-server route on `openai/*`.
**Gateway/startup:** Add owner-level startup trace attribution for auth, plugin loading, lookup counts, and plugin sidecar services.
**Channels/status reactions:** Wire `StatusReactionController` into WhatsApp message turns (queued → thinking → tool → done/error lifecycle, on par with Telegram and Discord), add `deploy`/`build`/`concierge` emoji categories with tool-token routing, and replace the status reaction defaults with self-explanatory emoji (🧠 thinking, 🛠️ tool, 💻 coding, 🌐 web, ⏳ stallSoft, ⚠️ stallHard, ✅ done, ❌ error, 🗜️ compacting) so stall and lifecycle reactions read as status indicators instead of emotional commentary. Fixes #59077.
**UI:** Add a browser-local Text size setting in Appearance and Quick Settings, scaling chat and dense UI text while keeping inputs above the mobile Safari focus-zoom threshold. Fixes #8547.
**Docs:** Add a dedicated ds4 provider page with local DeepSeek V4 Flash config, on-demand startup, context sizing, and live verification steps.
...and 12 more
🐛 Bug Fixes
**Cron/Codex:** Default exact-command scheduled agent turns to lightweight bootstrap context so automation runs the command before loading workspace identity or memory context.
**Codex plugin/Gateway:** Strip unpaired UTF-16 surrogates from Codex app-server JSON-RPC payloads and let stale reply-work recovery abort stalled reply runs, preventing malformed media turns from wedging gateway lanes.
**Codex app server:** Force OAuth refresh requests to perform a real token refresh instead of reusing unchanged inherited auth-profile tokens after refresh failures.
**UI:** Render `/tts audio` replies as playable audio attachments through the assistant-media ticket path, with structured-audio compatibility for older live payloads.
Bind gateway approval access to requester metadata [AI].
**Telegram:** Let isolated polling drain independent topics, DMs, and status/control commands concurrently while preserving same-lane order.
**UI:** Keep externalized Slack, OpenShell sandbox, and Anthropic Vertex runtime dependency declarations out of the root dist artifact build.
**ClawHub:** Include Amazon Bedrock and Bedrock Mantle provider packages in the published registry metadata so the externalized providers are discoverable from ClawHub as well as npm.
Auto-reply/Claude CLI: bridge CLI-runtime assistant text-delta agent events into the chat reasoning preview through `onReasoningStream`, mirroring the existing assistant-text and tool-event bridges and adding gating so non-CLI runtimes are unaffected. Thanks @anagnorisis2peripeteia and @pashpashpash.
**Mantis:** Keep QA evidence in Actions artifacts only and stop publishing evidence files to Git-backed artifact branches.
**Leaner installs:** WhatsApp, Slack, Amazon Bedrock, Anthropic Vertex, and related provider/plugin dependency cones moved out of the core runtime so installs only pull what you use.
**Telegram got much more resilient:** Isolated polling, durable local spooling, safer group-media handling, and preserved HTML/Markdown formatting in streamed and scheduled replies.
**Codex/OpenAI paths are smoother:** Auth-profile-backed media tools, MCP server projection, context-engine thread rotation, and better app-server/runtime fallback behavior.
Plugin installs and updates are harder to wedge, with pnpm 11 support, peer-dependency preservation, safer runtime scans, and source/git install fixes.
Gateway, browser, Slack, node pairing, sandbox, and transcript paths picked up a broad security/provenance hardening pass.
UI and reply delivery improved across Control UI, WebChat, TUI, rich-only replies, session history, and streaming auto-scroll.
**Amazon Bedrock:** Externalize the Bedrock and Bedrock Mantle provider packages so core installs no longer pull AWS SDK dependencies unless those providers are installed.
**Plugins:** Externalize Slack, OpenShell sandbox, and Anthropic Vertex so their runtime dependency cones install only when those plugins are installed.
**UI:** Add a persisted auto-scroll mode selector so users can keep the current near-bottom behavior, always follow streaming output, or turn automatic streaming scroll off and use the New messages button manually. Fixes #7648 and #81287.
**ACP:** Add `acp.fallbacks` so ACP turns can try configured backup runtime backends when the primary backend is unavailable before any output is emitted.
...and 72 more
🐛 Bug Fixes
**Doctor/Codex:** Stop warning that the message tool is unavailable for source-reply paths where OpenClaw grants `message` at runtime, keeping update and doctor output aligned with the OpenAI happy path.
**Channels/Weixin:** Bump the external Weixin catalog entry to `@tencent-weixin/openclaw-weixin@2.4.3` with the matching package integrity.
**Agents/subagents:** Apply `agents.defaults.subagents.model` before target agent primary models during `sessions_spawn`, so model-scoped runtimes such as `claude-cli` stay attached to default child runs. Fixes #81395.
**Telegram:** Keep Bot API polling alive during main event-loop stalls by moving ingress to an isolated worker with a durable local spool. Fixes #81132.
**Telegram:** Preserve rendered HTML formatting through lazy cron announce delivery so Markdown links stay clickable instead of falling back to literal anchor tags. Fixes #81742.
**Telegram:** Skip unmentioned group media before download when `requireMention` is active, avoiding failed media-download replies for messages that should be ignored. Fixes #81181.
**CLI:** Keep bare plugin and parent-command help on the lightweight path, avoiding plugin registry discovery before rendering help.
**Gateway/session history:** Carry monotonic transcript message sequence through live updates and refresh SSE history when stale sequence input would otherwise append bad incremental state.
**🔒 Security/sandbox:** Include Windows `USERPROFILE` in the sandbox blocked home roots so credential-bearing binds (such as `.codex`, `.openclaw`, or `.ssh` under the Windows user profile) are denied even when `HOME` points at a different shell home.
**🔒 Models config/auth:** Stop inferring provider env-var markers from broad `^[A-Z_][A-Z0-9_]*$` strings, and resolve config-backed provider `apiKey` values only through structured env SecretRefs (`secrets.providers[id]` / `secrets.defaults`), so unrelated env vars cannot accidentally become provider credentials.
**Amazon Bedrock:** Externalize the Bedrock and Bedrock Mantle provider packages so core installs no longer pull AWS SDK dependencies unless those providers are installed.
**Plugins:** Externalize Slack, OpenShell sandbox, and Anthropic Vertex so their runtime dependency cones install only when those plugins are installed.
**UI:** Add a persisted auto-scroll mode selector so users can keep the current near-bottom behavior, always follow streaming output, or turn automatic streaming scroll off and use the New messages button manually. Fixes #7648 and #81287.
**ACP:** Add `acp.fallbacks` so ACP turns can try configured backup runtime backends when the primary backend is unavailable before any output is emitted.
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside explicit SDK OAuth auth-result config patches, so provider helpers emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in per-agent config defaults and auth patches, so agent-specific emitted config keeps targeting `google/gemini-3.1-pro-preview`.
...and 66 more
🐛 Bug Fixes
**Agents/subagents:** Apply `agents.defaults.subagents.model` before target agent primary models during `sessions_spawn`, so model-scoped runtimes such as `claude-cli` stay attached to default child runs. Fixes #81395.
**Telegram:** Keep Bot API polling alive during main event-loop stalls by moving ingress to an isolated worker with a durable local spool. Fixes #81132.
**Telegram:** Preserve rendered HTML formatting through lazy cron announce delivery so Markdown links stay clickable instead of falling back to literal anchor tags. Fixes #81742.
**Telegram:** Skip unmentioned group media before download when `requireMention` is active, avoiding failed media-download replies for messages that should be ignored. Fixes #81181.
**CLI:** Keep bare plugin and parent-command help on the lightweight path, avoiding plugin registry discovery before rendering help.
**Gateway/session history:** Carry monotonic transcript message sequence through live updates and refresh SSE history when stale sequence input would otherwise append bad incremental state.
**🔒 Security/sandbox:** Include Windows `USERPROFILE` in the sandbox blocked home roots so credential-bearing binds (such as `.codex`, `.openclaw`, or `.ssh` under the Windows user profile) are denied even when `HOME` points at a different shell home.
**🔒 Models config/auth:** Stop inferring provider env-var markers from broad `^[A-Z_][A-Z0-9_]*$` strings, and resolve config-backed provider `apiKey` values only through structured env SecretRefs (`secrets.providers[id]` / `secrets.defaults`), so unrelated env vars cannot accidentally become provider credentials.
**Media fetch:** Skip allocating and buffering the response body for bodyless media responses (HEAD probes and 204-style empty bodies), avoiding wasted heap on streams that carry no payload.
**CLI:** Forward provider-specific auth flags (e.g. `--openai-api-key`) through the onboarding wizard so they reach provider auth methods via `ctx.opts`, letting `--openai-api-key "$OPENAI_API_KEY"` skip the redundant "use existing env var?" prompt in non-interactive harnesses.
**Amazon Bedrock:** Externalize the Bedrock and Bedrock Mantle provider packages so core installs no longer pull AWS SDK dependencies unless those providers are installed.
**Plugins:** Externalize Slack, OpenShell sandbox, and Anthropic Vertex so their runtime dependency cones install only when those plugins are installed.
**UI:** Add a persisted auto-scroll mode selector so users can keep the current near-bottom behavior, always follow streaming output, or turn automatic streaming scroll off and use the New messages button manually. Fixes #7648 and #81287.
**ACP:** Add `acp.fallbacks` so ACP turns can try configured backup runtime backends when the primary backend is unavailable before any output is emitted.
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside explicit SDK OAuth auth-result config patches, so provider helpers emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in per-agent config defaults and auth patches, so agent-specific emitted config keeps targeting `google/gemini-3.1-pro-preview`.
...and 66 more
🐛 Bug Fixes
**Telegram:** Keep Bot API polling alive during main event-loop stalls by moving ingress to an isolated worker with a durable local spool. Fixes #81132.
**CLI:** Keep bare plugin and parent-command help on the lightweight path, avoiding plugin registry discovery before rendering help.
**Gateway/session history:** Carry monotonic transcript message sequence through live updates and refresh SSE history when stale sequence input would otherwise append bad incremental state.
**🔒 Security/sandbox:** Include Windows `USERPROFILE` in the sandbox blocked home roots so credential-bearing binds (such as `.codex`, `.openclaw`, or `.ssh` under the Windows user profile) are denied even when `HOME` points at a different shell home.
**🔒 Models config/auth:** Stop inferring provider env-var markers from broad `^[A-Z_][A-Z0-9_]*$` strings, and resolve config-backed provider `apiKey` values only through structured env SecretRefs (`secrets.providers[id]` / `secrets.defaults`), so unrelated env vars cannot accidentally become provider credentials.
**Media fetch:** Skip allocating and buffering the response body for bodyless media responses (HEAD probes and 204-style empty bodies), avoiding wasted heap on streams that carry no payload.
**CLI:** Forward provider-specific auth flags (e.g. `--openai-api-key`) through the onboarding wizard so they reach provider auth methods via `ctx.opts`, letting `--openai-api-key "$OPENAI_API_KEY"` skip the redundant "use existing env var?" prompt in non-interactive harnesses.
**CLI:** Drop trailing periods from Codex migrate item messages and `REASON_CODE_MESSAGES` strings so plan/result rows read as labels instead of sentence fragments.
**Slack:** Treat malformed private-file redirect `Location` headers as unfollowable redirects instead of failing Slack media downloads.
**Plugins:** Discover provider plugins from `setup.providers[].envVars` credentials during provider discovery while keeping the deprecated `providerAuthEnvVars` fallback.
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside explicit SDK OAuth auth-result config patches, so provider helpers emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in per-agent config defaults and auth patches, so agent-specific emitted config keeps targeting `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in provider catalog rows when API-key onboarding only reapplies the agent default, so emitted config keeps testing `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in `config set` mutation output for agent overrides and provider catalog rows, so current config emits `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Canonicalize provider-qualified retired Gemini 3 Pro Preview refs during Google forward-compatible model resolution, so emitted config uses `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize proxy-prefixed retired Gemini 3 Pro Preview catalog rows, so emitted configs use `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
...and 62 more
🐛 Bug Fixes
**iMessage:** Stop sending visible `` placeholder text for media-only native image sends while preserving the internal echo key that prevents self-echo duplicate replies.
**Agents/sessions:** Create configured agent main sessions before first `sessions_send` or gateway send, so agent-to-agent messages no longer fail when the target agent has not started yet.
**gateway:** Pass Talk session scope to resolver [AI].
**Gateway protocol:** Require v4 clients and stream explicit chat `deltaText`/`replace` frames so SDK clients can consume assistant updates without local diffing.
**GitHub Copilot:** Exchange OAuth tokens for Copilot API tokens on image understanding requests and route Gemini image payloads through Chat Completions, fixing Copilot Gemini image descriptions.
**Gateway:** Hide pending Node pairing commands, capabilities, and permissions until approval, and refresh the live approved surface when pairings change.
**Plugins/Feishu/WhatsApp/Line:** Enforce inbound media size caps while reading download streams, avoiding full buffering of oversized attachments.
**Plugins/install:** Limit install-time code safety scans to plugin-owned runtime entrypoints while keeping dependency manifest denylist checks, so trusted packages with large dependency trees no longer get blocked or warned on third-party runtime internals.
**Config:** Serialize and retry semantic config mutations centrally, so concurrent commands can rebase safe changes instead of clobbering or hand-rolling command-local retry loops.
Require approval for setup-code device pairing [AI].
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside explicit SDK OAuth auth-result config patches, so provider helpers emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in per-agent config defaults and auth patches, so agent-specific emitted config keeps targeting `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in provider catalog rows when API-key onboarding only reapplies the agent default, so emitted config keeps testing `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in `config set` mutation output for agent overrides and provider catalog rows, so current config emits `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Canonicalize provider-qualified retired Gemini 3 Pro Preview refs during Google forward-compatible model resolution, so emitted config uses `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize proxy-prefixed retired Gemini 3 Pro Preview catalog rows, so emitted configs use `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
...and 62 more
🐛 Bug Fixes
**gateway:** Pass Talk session scope to resolver [AI].
**Gateway protocol:** Require v4 clients and stream explicit chat `deltaText`/`replace` frames so SDK clients can consume assistant updates without local diffing.
**GitHub Copilot:** Exchange OAuth tokens for Copilot API tokens on image understanding requests and route Gemini image payloads through Chat Completions, fixing Copilot Gemini image descriptions.
**Gateway:** Hide pending Node pairing commands, capabilities, and permissions until approval, and refresh the live approved surface when pairings change.
**Plugins/Feishu/WhatsApp/Line:** Enforce inbound media size caps while reading download streams, avoiding full buffering of oversized attachments.
**Config:** Serialize and retry semantic config mutations centrally, so concurrent commands can rebase safe changes instead of clobbering or hand-rolling command-local retry loops.
Require approval for setup-code device pairing [AI].
**Plugins/install:** Preserve third-party peer dependencies in the managed npm root when later plugin installs or updates recalculate the shared dependency tree.
**Plugins/uninstall:** Prune managed third-party peer dependencies after their owning npm plugin is removed, without blocking plugin cleanup on peer-prune failures.
**Docker:** Pin setup-time container paths so stale host `.env` OpenClaw paths cannot leak into Linux containers. Fixes #80381.
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside explicit SDK OAuth auth-result config patches, so provider helpers emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in provider catalog rows when API-key onboarding only reapplies the agent default, so emitted config keeps testing `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Canonicalize provider-qualified retired Gemini 3 Pro Preview refs during Google forward-compatible model resolution, so emitted config uses `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize proxy-prefixed retired Gemini 3 Pro Preview catalog rows, so emitted configs use `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside per-agent model overrides before writing config, so agent-specific config emits `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Docs/subagents:** Document `agents.defaults.subagents.announceTimeoutMs` in the sub-agent and configuration references.
...and 59 more
🐛 Bug Fixes
**Codex runtime:** Allow the official installed `@openclaw/codex` package to use its private task-runtime SDK helper, fixing `MODULE_NOT_FOUND` during migrated OpenAI/Codex beta runs.
**Codex migration:** Make Enter activate the highlighted checkbox row before continuing, so `Skip for now` and bulk-selection rows work even when planned items start preselected.
**Codex harness:** Keep auth-profile-backed media tools such as `image_generate` available when OpenAI auth lives in the agent's auth-profile store instead of environment variables.
**WhatsApp/install:** Allow Baileys' pinned libsignal git subdependency under pnpm 11 so source installs and local checks can complete.
Fix(memory-wiki): require admin scope for ingest [AI].
Memory-wiki: require write scope for Obsidian search [AI].
**UI:** Skip copied metadata for bundled plugins that are excluded from build entries, preventing update/status rebuilds from advertising missing QQ Bot runtime files.
**UI:** Nest subagent sessions under their parent session in the session picker dropdown using a visual `└─ ` prefix, making the parent-child relationship clear. Fixes #77628.
Auto-reply: surface a visible error when the configured model backend fails and fallback produces no visible reply, while preserving intentional silent turns and side-effect-only deliveries.
**Agents/exec:** Skip redundant heartbeat wake-ups for subagent session exec completions, preventing spurious LLM invocations on parent sessions. Fixes #66748.
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in provider catalog rows when API-key onboarding only reapplies the agent default, so emitted config keeps testing `google/gemini-3.1-pro-preview`.
**Docs/subagents:** Document `agents.defaults.subagents.announceTimeoutMs` in the sub-agent and configuration references.
**Cron:** Add direct `cron.get`, `openclaw cron get `, and agent-tool `get` support for inspecting one stored cron job by id.
**Agents/tools:** Add per-sender tool policies with canonical channel-scoped sender keys, so operators can restrict dangerous tools by requester identity across global, agent, group, core, bundled, and plugin tool surfaces.
**ACP:** Expose Gateway session lineage metadata through ACP session listings and session info snapshots so clients can render subagent graphs without private Gateway side channels.
**Channels/iMessage:** Add `openclaw channels status --channel ` filtering and document the BlueBubbles-to-imsg cutover path so operators can probe iMessage without starting both channel monitors.
...and 55 more
🐛 Bug Fixes
**Codex harness:** Keep auth-profile-backed media tools such as `image_generate` available when OpenAI auth lives in the agent's auth-profile store instead of environment variables.
**WhatsApp/install:** Allow Baileys' pinned libsignal git subdependency under pnpm 11 so source installs and local checks can complete.
**Codex harness:** Keep auth-profile-backed media tools such as `image_generate` available when OpenAI auth lives in the agent's auth-profile store instead of environment variables.
Fix(memory-wiki): require admin scope for ingest [AI].
Memory-wiki: require write scope for Obsidian search [AI].
**UI:** Skip copied metadata for bundled plugins that are excluded from build entries, preventing update/status rebuilds from advertising missing QQ Bot runtime files.
**UI:** Nest subagent sessions under their parent session in the session picker dropdown using a visual `└─ ` prefix, making the parent-child relationship clear. Fixes #77628.
Auto-reply: surface a visible error when the configured model backend fails and fallback produces no visible reply, while preserving intentional silent turns and side-effect-only deliveries.
**Agents/exec:** Skip redundant heartbeat wake-ups for subagent session exec completions, preventing spurious LLM invocations on parent sessions. Fixes #66748.
**Provider streams:** Keep OpenAI-compatible SSE and JSON fallback streams draining across split chunks and fail Azure Responses streams with a bounded first-event diagnostic instead of stalling. Refs #80926. Thanks @galiniliev and @CaptainTimon.
**Gateway/OpenAI HTTP:** Honor `max_completion_tokens` and `max_tokens` on inbound `/v1/chat/completions` requests so client-provided token caps reach the upstream provider via `streamParams.maxTokens`, with `max_completion_tokens` taking precedence when both are sent.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in provider catalog rows when API-key onboarding only reapplies the agent default, so emitted config keeps testing `google/gemini-3.1-pro-preview`.
**Docs/subagents:** Document `agents.defaults.subagents.announceTimeoutMs` in the sub-agent and configuration references.
**Cron:** Add direct `cron.get`, `openclaw cron get `, and agent-tool `get` support for inspecting one stored cron job by id.
**Agents/tools:** Add per-sender tool policies with canonical channel-scoped sender keys, so operators can restrict dangerous tools by requester identity across global, agent, group, core, bundled, and plugin tool surfaces.
**ACP:** Expose Gateway session lineage metadata through ACP session listings and session info snapshots so clients can render subagent graphs without private Gateway side channels.
**Channels/iMessage:** Add `openclaw channels status --channel ` filtering and document the BlueBubbles-to-imsg cutover path so operators can probe iMessage without starting both channel monitors.
...and 55 more
🐛 Bug Fixes
**Codex harness:** Keep auth-profile-backed media tools such as `image_generate` available when OpenAI auth lives in the agent's auth-profile store instead of environment variables.
**WhatsApp/install:** Allow Baileys' pinned libsignal git subdependency under pnpm 11 so source installs and local checks can complete.
**Codex harness:** Keep auth-profile-backed media tools such as `image_generate` available when OpenAI auth lives in the agent's auth-profile store instead of environment variables.
Fix(memory-wiki): require admin scope for ingest [AI].
Memory-wiki: require write scope for Obsidian search [AI].
**UI:** Skip copied metadata for bundled plugins that are excluded from build entries, preventing update/status rebuilds from advertising missing QQ Bot runtime files.
**UI:** Nest subagent sessions under their parent session in the session picker dropdown using a visual `└─ ` prefix, making the parent-child relationship clear. Fixes #77628.
Auto-reply: surface a visible error when the configured model backend fails and fallback produces no visible reply, while preserving intentional silent turns and side-effect-only deliveries.
**Agents/exec:** Skip redundant heartbeat wake-ups for subagent session exec completions, preventing spurious LLM invocations on parent sessions. Fixes #66748.
**Provider streams:** Keep OpenAI-compatible SSE and JSON fallback streams draining across split chunks and fail Azure Responses streams with a bounded first-event diagnostic instead of stalling. Refs #80926. Thanks @galiniliev and @CaptainTimon.
**🔒 Models/OpenAI CLI auth:** Make `openclaw models auth login --provider openai` start the ChatGPT/Codex account login by default, while `--method api-key` remains the explicit OpenAI API-key setup path.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids inside SDK OAuth auth-result default config patches, so helper-built provider auth flows emit `google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids returned by direct `openclaw models auth login --set-default` provider auth flows before writing config, so Gemini testing targets `google/gemini-3.1-pro-preview`.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview ids in provider catalog rows when API-key onboarding only reapplies the agent default, so emitted config keeps testing `google/gemini-3.1-pro-preview`.
**Docs/subagents:** Document `agents.defaults.subagents.announceTimeoutMs` in the sub-agent and configuration references.
**Cron:** Add direct `cron.get`, `openclaw cron get `, and agent-tool `get` support for inspecting one stored cron job by id.
**Agents/tools:** Add per-sender tool policies with canonical channel-scoped sender keys, so operators can restrict dangerous tools by requester identity across global, agent, group, core, bundled, and plugin tool surfaces.
**ACP:** Expose Gateway session lineage metadata through ACP session listings and session info snapshots so clients can render subagent graphs without private Gateway side channels.
**Channels/iMessage:** Add `openclaw channels status --channel ` filtering and document the BlueBubbles-to-imsg cutover path so operators can probe iMessage without starting both channel monitors.
**CI:** Add a non-blocking `plugin-inspector-advisory` artifact to Plugin Prerelease so release runs capture bundled plugin compatibility triage without changing the blocking gate.
...and 54 more
🐛 Bug Fixes
Fix(memory-wiki): require admin scope for ingest [AI].
Memory-wiki: require write scope for Obsidian search [AI].
**UI:** Skip copied metadata for bundled plugins that are excluded from build entries, preventing update/status rebuilds from advertising missing QQ Bot runtime files.
**UI:** Nest subagent sessions under their parent session in the session picker dropdown using a visual `└─ ` prefix, making the parent-child relationship clear. Fixes #77628.
**CLI:** Render terminal QR codes with full-block characters by default so the bundled `qrcode` terminal renderer does not emit a pathologically dense ANSI final row in compact half-block mode that breaks scanning in some terminals. Fixes #77820.
**Agents/compaction:** Read post-compaction AGENTS.md refresh context from the queued run workspace instead of the runner process cwd, so CLI-backed follow-up turns re-inject the correct workspace startup rules after compaction. Fixes #70541.
**Agents/read tool:** Treat positive offsets beyond EOF as empty ranges instead of surfacing the upstream read error, so stale pagination cursors no longer crash tool calls while unrelated read failures still fail loud. Fixes #62466.
**Google/Gemini:** Normalize retired Gemini 3 Pro Preview refs left in Google API-key onboarding model allowlists and fallbacks, so setup-emitted config keeps testing `google/gemini-3.1-pro-preview` instead of `google/gemini-3-pro-preview`.
**Google/Gemini:** Normalize retired nested Gemini 3 Pro Preview ids when resolving exact configured proxy-provider refs, so `kilocode/google/gemini-3-pro-preview` resolves to `kilocode/google/gemini-3.1-pro-preview` for Gemini 3.1 testing.
**CLI:** Strip generic OSC terminal escape payloads from sanitized output fields, preventing clipboard/title escape bodies from leaking into commitment tables and other terminal-safe text.
**CI:** Add a non-blocking `plugin-inspector-advisory` artifact to Plugin Prerelease so release runs capture bundled plugin compatibility triage without changing the blocking gate.
**Runtime/Fly:** Detect Fly Machines as container environments from their runtime env vars, so gateway bind and Bonjour defaults match remote container launches.
**Providers/fal:** Route GPT Image 2 and Nano Banana 2 reference-image edit requests to `/edit` with `image_urls` array, enforce NB2 edit geometry using `aspect_ratio` and `resolution` params, lift Fal edit mode input-image caps to 10 for GPT Image 2 and 14 for Nano Banana 2, and allow aspect-ratio hints in edit mode.
**UI:** Show a plain HTML recovery panel when the app module never registers, giving blank dashboard pages a retry path and browser-extension troubleshooting link. Fixes #44107.
**UI:** Enable additional low-churn oxlint rules for promise, TypeScript, and runtime footgun checks.
**UI:** Enable stricter Vitest lint rules for focused, disabled, conditional, hook, matcher, and expectation hazards.
**UI:** Pin explicit oxfmt defaults in the shared formatter config to keep formatting behavior stable across upgrades.
**TypeScript:** Enable stricter compiler checks for implicit returns, side-effect imports, overrides, and unused production code.
**Logging:** Add targeted model transport, payload, SSE, and code-mode diagnostics with redacted URL handling.
**Agents:** Allow `session.agentToAgent.maxPingPongTurns` up to 20 while keeping the default at 5 for longer agent-to-agent reply chains. Fixes #52382.
...and 39 more
🐛 Bug Fixes
**Bonjour/Gateway:** Treat active ciao probing and fresh name-conflict renames as in-progress so the mDNS watchdog waits for probe settlement before retrying, preventing rapid re-advertise loops on Windows, WSL, and other multicast-hostile hosts. Refs #74242.
**Providers/MiniMax:** Send a minimal Anthropic-compatible user fallback when message conversion filters a turn to an empty payload, so MiniMax M2.7 no longer returns `chat content is empty` after tool-heavy sessions. Fixes #74589. Thanks @neeravmakwana and @DerekEXS.
**Cron:** Keep long manual cron runs active in the task registry until completion, preventing transient `lost` markers before durable recovery reconciles. Fixes #78233.
**CLI:** Surface a `GH_CONFIG_DIR` hint when the GitHub skill is usable but `gh` auth lives under a different operator HOME than the agent process, without warning for disabled or filtered skills. Fixes #78063.
**Gateway:** Dedupe concurrent `send`, `poll`, and `message.action` requests while delivery is still in flight, preventing duplicate outbound work for the same idempotency key.
**Cron:** Keep main-session `systemEvent` heartbeat wakes on their bound session route for both direct and queued wake paths by dropping inherited explicit heartbeat destinations when forcing `target: "last"`. Fixes #73900.
**Telegram:** Honor forced document delivery for video media so `--force-document` sends MP4s as documents instead of typed videos. Fixes #80389.
**Gateway:** Clear speculative node wake state when APNs registration is missing, preventing unregistered or mistyped node IDs from retaining wake throttle entries. Fixes #68847.
Auto-reply: keep late follow-up queue drain finalizers from deleting a replacement queue registered after `/stop`, preventing immediate follow-up messages from being orphaned. Fixes #68838.
**Feishu:** Make manual App ID/App Secret setup the default channel-binding path while keeping QR scan-to-create as an optional best-effort flow, and document the manual fallback for domestic Feishu mobile clients that do not react to the QR code. Fixes #80591.
**CI:** Add a non-blocking `plugin-inspector-advisory` artifact to Plugin Prerelease so release runs capture bundled plugin compatibility triage without changing the blocking gate.
**Runtime/Fly:** Detect Fly Machines as container environments from their runtime env vars, so gateway bind and Bonjour defaults match remote container launches.
**Providers/fal:** Route GPT Image 2 and Nano Banana 2 reference-image edit requests to `/edit` with `image_urls` array, enforce NB2 edit geometry using `aspect_ratio` and `resolution` params, lift Fal edit mode input-image caps to 10 for GPT Image 2 and 14 for Nano Banana 2, and allow aspect-ratio hints in edit mode.
**UI:** Show a plain HTML recovery panel when the app module never registers, giving blank dashboard pages a retry path and browser-extension troubleshooting link. Fixes #44107.
**UI:** Enable additional low-churn oxlint rules for promise, TypeScript, and runtime footgun checks.
**UI:** Enable stricter Vitest lint rules for focused, disabled, conditional, hook, matcher, and expectation hazards.
**UI:** Pin explicit oxfmt defaults in the shared formatter config to keep formatting behavior stable across upgrades.
**TypeScript:** Enable stricter compiler checks for implicit returns, side-effect imports, overrides, and unused production code.
**Logging:** Add targeted model transport, payload, SSE, and code-mode diagnostics with redacted URL handling.
**Agents:** Allow `session.agentToAgent.maxPingPongTurns` up to 20 while keeping the default at 5 for longer agent-to-agent reply chains. Fixes #52382.
...and 39 more
🐛 Bug Fixes
**Bonjour/Gateway:** Treat active ciao probing and fresh name-conflict renames as in-progress so the mDNS watchdog waits for probe settlement before retrying, preventing rapid re-advertise loops on Windows, WSL, and other multicast-hostile hosts. Refs #74242.
**Providers/MiniMax:** Send a minimal Anthropic-compatible user fallback when message conversion filters a turn to an empty payload, so MiniMax M2.7 no longer returns `chat content is empty` after tool-heavy sessions. Fixes #74589. Thanks @neeravmakwana and @DerekEXS.
**Cron:** Keep long manual cron runs active in the task registry until completion, preventing transient `lost` markers before durable recovery reconciles. Fixes #78233.
**CLI:** Surface a `GH_CONFIG_DIR` hint when the GitHub skill is usable but `gh` auth lives under a different operator HOME than the agent process, without warning for disabled or filtered skills. Fixes #78063.
**Gateway:** Dedupe concurrent `send`, `poll`, and `message.action` requests while delivery is still in flight, preventing duplicate outbound work for the same idempotency key.
**Cron:** Keep main-session `systemEvent` heartbeat wakes on their bound session route for both direct and queued wake paths by dropping inherited explicit heartbeat destinations when forcing `target: "last"`. Fixes #73900.
**Telegram:** Honor forced document delivery for video media so `--force-document` sends MP4s as documents instead of typed videos. Fixes #80389.
**Gateway:** Clear speculative node wake state when APNs registration is missing, preventing unregistered or mistyped node IDs from retaining wake throttle entries. Fixes #68847.
Auto-reply: keep late follow-up queue drain finalizers from deleting a replacement queue registered after `/stop`, preventing immediate follow-up messages from being orphaned. Fixes #68838.
**Feishu:** Make manual App ID/App Secret setup the default channel-binding path while keeping QR scan-to-create as an optional best-effort flow, and document the manual fallback for domestic Feishu mobile clients that do not react to the QR code. Fixes #80591.
**UI:** Enable stricter Vitest lint rules for focused, disabled, conditional, hook, matcher, and expectation hazards.
**UI:** Pin explicit oxfmt defaults in the shared formatter config to keep formatting behavior stable across upgrades.
**TypeScript:** Enable stricter compiler checks for implicit returns, side-effect imports, overrides, and unused production code.
**UI:** Upgrade workspace package management to pnpm 11 and keep Docker, install, update, and release workflows on the pnpm 11 config surface.
**Models:** Add provider-level `localService` startup for on-demand local model servers before OpenAI-compatible requests, including one-shot model probes.
**Agents:** Trim default system prompt guidance and send-only message tool schemas to reduce prompt tokens while preserving GPT-5 personality guidance.
**Context:** Add `/context map` to send a treemap image of the current session context contributors.
**Slack:** Add `unfurlLinks` and `unfurlMedia` config for bot `chat.postMessage` replies, including per-account overrides, so Slack link and media previews can be suppressed without workspace-wide settings. Fixes #48435. Thanks @esegev1 and @HemantSudarshan.
**Slack:** Add explicit `replyBroadcast` support for text and Block Kit thread replies so agents can opt into Slack's parent-channel `reply_broadcast` behavior.
**Slack:** Preserve mention target/source metadata in inbound prompt context so agents can distinguish direct bot mentions from implicit thread wakes that mention someone else. Fixes #79025.
...and 23 more
🐛 Bug Fixes
**Google/Gemini:** Normalize retired nested Gemini 3 Pro Preview ids while converting manifest catalog rows into emitted provider config, so `google/gemini-3.1-pro-preview` is used for testing instead of `google/gemini-3-pro-preview`.
**Native apps:** Advertise the Gateway protocol compatibility range so chat and node sessions can connect to v3 gateways after additive v4 client updates.
**Gateway/agents:** Keep stale `sessions_send` ACP manager and `web_fetch` runtime chunks importable after package updates, preventing live gateways from breaking before restart. Fixes #78804.
**Gateway:** Avoid synchronous restart-sentinel state probes during post-attach startup, preventing slow Windows or redirected state directories from blocking channel turns. Fixes #79264.
**🔒 Agents/auth:** Update successful model auth profile status with one locked store write, reducing post-model reply latency from duplicate `auth-profiles.json` saves.
**Agents/image:** Honor explicit `image` tool model overrides even when `agents.defaults.imageModel` is unset, restoring one-off vision calls for configured multimodal providers. Fixes #79341.
Codex app-server: preserve prompt-local current-turn context through context-engine prompt projection, so replied-to Telegram messages stay visible to the Codex model input.
**Telegram:** Pass agent-scoped media roots through gateway message actions so workspace-local media from the active agent is not rejected as cross-agent access.
**CLI:** Keep `gateway status --deep` plugin-aware so configured plugin manifest warnings, including missing channel config metadata, stay visible during install and update smoke checks.
**Feishu:** Fall back to a top-level group send when normal group quoted replies target a withdrawn or missing message, preventing replies from disappearing silently while preserving native topic safety. Fixes #79349.
**QA/Mantis:** Add Telegram live PR evidence automation with Convex-leased credentials, Crabbox transcript capture, motion GIF previews, and inline PR comments.
**QA/Mantis:** Add a Telegram desktop scenario builder that leases Crabbox, installs native Telegram Desktop, configures an OpenClaw Telegram gateway with leased bot credentials, and records VNC screenshot/video artifacts.
**Discord/voice:** Add realtime voice diagnostics for speaker turns, playback resets, barge-in detection, and audio cutoff analysis.
**Talk:** Add `talk.realtime.instructions` so operators can append realtime voice style instructions while preserving OpenClaw's built-in agent-consult guidance.
**Discord/voice:** Default test and source installs to the pure-JS `opusscript` decoder by ignoring optional native `@discordjs/opus` builds, avoiding slow native addon compiles outside dedicated voice-performance lanes.
**Discord/voice:** Add an opt-in native `@discordjs/opus` install script and decoder preference for live voice-performance lanes without charging unrelated Docker/tests for native addon builds.
**Gateway/skills:** Add an opt-in private skill archive upload install path gated by `skills.install.allowUploadedArchives`, so trusted Gateway clients can stage and install zip-backed skills only when operators explicitly enable the code-install surface.
**Dependencies:** Refresh workspace pins and patch targets, including ACPX `@agentclientprotocol/claude-agent-acp` `0.33.1`, Codex ACP `0.14.0`, Baileys `7.0.0-rc10`, Google GenAI `2.0.1`, OpenAI `6.37.0`, AWS SDK `3.1045.0`, Kysely `0.29.0`, Tlon skill `0.3.6`, Aimock `1.19.5`, and tsdown `0.22.0`.
**Agents/compaction:** Preserve scoped background exec/process session references across embedded compaction and after-turn runtime contexts without exposing sessions from unrelated scopes. Fixes #79284.
**CLI:** Improve setup, onboarding, configure, and channel command wayfinding so terminal flows explain the next useful command instead of relying on terse setup labels.
...and 1 more
🐛 Bug Fixes
**Telegram:** Pass agent-scoped media roots through gateway message actions so workspace-local media from the active agent is not rejected as cross-agent access.
**CLI:** Keep `gateway status --deep` plugin-aware so configured plugin manifest warnings, including missing channel config metadata, stay visible during install and update smoke checks.
**ACPX:** Run and await the embedded ACP backend startup probe by default so the gateway `ready` signal no longer fires before the acpx runtime has either become usable or reported a probe failure; set `OPENCLAW_ACPX_RUNTIME_STARTUP_PROBE=0` to restore lazy startup. Fixes #79596.
OpenAI-compatible models: strip prior assistant reasoning fields from replayed Chat Completions history by default, preventing oMLX/vLLM Qwen follow-up turns from rejecting or stalling on stale `reasoning` payloads. Fixes #46637. Thanks @zipzagster and @lexhoefsloot.
**CLI:** Give non-Azure custom providers a safe generated context window and heal legacy 4k wizard entries without overwriting explicit valid small model limits, preventing first-turn compaction loops. Fixes #79428.
OpenAI-compatible models: **`compat.strictMessageKeys` to strip Chat Completions replay messages to `role` and `content` for strict** is now supported as a providers that reject OpenAI-style tool and metadata keys. Fixes #50374.
**Voice/Ollama:** Honor routed voice agent `tools.allow` for classic embedded voice responses, including empty allowlists, so no-tool Ollama agents do not receive tool schemas. Fixes #79506.
**Gateway:** Reread config from disk after the first in-process restart loop startup, preventing SIGUSR1 restarts from reusing a stale startup snapshot and dropping config written after boot. Fixes #79947.
**Agents:** Apply the LLM idle watchdog while provider stream setup is still pending, preventing silent pre-stream model hangs from waiting for the full agent timeout.
**Cron:** Let isolated self-cleanup runs inspect their own job run history while keeping other cron jobs and mutation actions blocked. Fixes #80019.
**QA/Mantis:** Add Telegram live PR evidence automation with Convex-leased credentials, Crabbox transcript capture, motion GIF previews, and inline PR comments.
**QA/Mantis:** Add a Telegram desktop scenario builder that leases Crabbox, installs native Telegram Desktop, configures an OpenClaw Telegram gateway with leased bot credentials, and records VNC screenshot/video artifacts.
**Discord/voice:** Add realtime voice diagnostics for speaker turns, playback resets, barge-in detection, and audio cutoff analysis.
**Talk:** Add `talk.realtime.instructions` so operators can append realtime voice style instructions while preserving OpenClaw's built-in agent-consult guidance.
**Discord/voice:** Default test and source installs to the pure-JS `opusscript` decoder by ignoring optional native `@discordjs/opus` builds, avoiding slow native addon compiles outside dedicated voice-performance lanes.
**Gateway/skills:** Add an opt-in private skill archive upload install path gated by `skills.install.allowUploadedArchives`, so trusted Gateway clients can stage and install zip-backed skills only when operators explicitly enable the code-install surface.
**Dependencies:** Refresh workspace pins and patch targets, including ACPX `@agentclientprotocol/claude-agent-acp` `0.33.1`, Codex ACP `0.14.0`, Baileys `7.0.0-rc10`, Google GenAI `2.0.1`, OpenAI `6.37.0`, AWS SDK `3.1045.0`, Kysely `0.29.0`, Tlon skill `0.3.6`, Aimock `1.19.5`, and tsdown `0.22.0`.
**Agents/compaction:** Preserve scoped background exec/process session references across embedded compaction and after-turn runtime contexts without exposing sessions from unrelated scopes. Fixes #79284.
**CLI:** Improve setup, onboarding, configure, and channel command wayfinding so terminal flows explain the next useful command instead of relying on terse setup labels.
**Agents/Codex:** Remove the configurable Codex dynamic-tools profile so Codex app-server always owns workspace, edit, patch, exec, process, and plan tools while OpenClaw integration tools remain available.
🐛 Bug Fixes
**Agents:** Apply the LLM idle watchdog while provider stream setup is still pending, preventing silent pre-stream model hangs from waiting for the full agent timeout.
**Cron:** Let isolated self-cleanup runs inspect their own job run history while keeping other cron jobs and mutation actions blocked. Fixes #80019.
**CLI:** Persist explicit `config set` and `config patch` values that equal runtime defaults instead of reporting success while dropping them. Fixes #79856. Thanks @abodanty and @hclsys.
**OpenAI/realtime voice:** Accept Codex-compatible legacy audio and transcript event aliases so provider protocol drift does not drop assistant audio or captions.
**Discord/voice:** Keep default agent-proxy realtime sessions from auto-speaking filler before the forced OpenClaw consult answer, finish Discord playback on realtime response completion, and queue later exact-speech answers until playback idles to avoid mid-sentence replacement.
**Gateway:** Return deterministic `400 invalid_request_error` responses for malformed encoded session-kill HTTP paths instead of letting route-shaped requests fall through to later Gateway handlers.
**UI:** Serve root PWA and favicon assets from `/__openclaw__/` SPA routes so tab icons, install metadata, and the service worker do not 404 after internal navigation. Fixes #80072.
**OpenAI/realtime voice:** Honor disabled input-audio interruption locally so server VAD speech-start events do not clear Discord playback after operators set `interruptResponseOnInputAudio: false`.
**Telegram:** Keep no-response DM turns quiet instead of rewriting them into visible silent-reply chatter. Fixes #78188.
**Telegram:** Handle managed select button callbacks before the raw callback fallback while preserving delimiter-containing option values such as `env|prod`.
**Channels/iMessage:** Remove the bundled BlueBubbles channel surface and deprecate BlueBubbles-backed iMessage setup in OpenClaw. Existing `channels.bluebubbles` configs must migrate to `channels.imessage` using `imsg` on a signed-in Mac or an SSH wrapper, and non-macOS default `imsg` configs now report remote-Mac wrapper guidance.
🆕 New Features
**CLI:** Add `/think default` and `/fast default` to clear session overrides and inherit configured/provider defaults.
**Dependencies:** Refresh workspace dependency pins and lockfile, including `@openai/codex` `0.130.0`, `acpx` `0.7.0`, AWS SDK `3.1044.0`, OpenTelemetry `0.217.0`, `typebox` `1.1.38`, `vite` `8.0.11`, `oxfmt` `0.48.0`, and `oxlint` `1.63.0`, and update the Codex harness model snapshot for the new bundled app-server catalog.
**Plugins/install:** Add guarded plugin install overrides so onboarding and repair tests can route specific plugins to registry specs or local `npm pack` artifacts via environment variables.
**Tests/Docker:** Add Codex on-demand install and live plugin-tool dependency E2E lanes for packaged onboarding and npm-pack plugin proof.
**Plugins/ACPX:** Accept an optional `args` array in `agents.` config so paths and flag values containing spaces stay intact when spawning ACP agent processes. Thanks @TheArchitectit and @BunsDev.
**Agents:** Inject the current provider/model identity into system prompts, including configured prompt overrides and CLI hook prompt replacements, so agents can answer model-identity questions from the actual runtime selection.
**CLI:** Add the optional bundled `oc-path` plugin, providing `openclaw path` for surgical `oc://` access to markdown, JSONC, and JSONL workspace files.
**Plugins/SDK:** Add unified model catalog registration for text, image, video, and music providers, including `providerCatalogEntry` manifests, shared media list help, live catalog caching, and per-model video capability overlays.
**Plugin SDK:** Add presentation helpers for controls-only interactive rendering and opt-in empty fallback text so rich channel renderers can share `MessagePresentation` semantics without duplicating native cards or components.
**CLI:** Make parser, startup, config, guardrail, channel, agent, task, session, and MCP failures explain what happened and point to the next recovery command.
...and 134 more
🐛 Bug Fixes
**CLI:** Handle resumed CLI JSONL output and bound supervisor output buffering so resumed runs stay readable without letting noisy child output grow unbounded.
**Agents/sandbox:** Include the container workspace path hint in sandbox-root escape errors while preserving shortened host workspace roots. Fixes #79712. Thanks @haumanto and @hclsys.
**Image generation:** Honor configured web-fetch SSRF policy across OpenAI, Google, MiniMax, OpenRouter, and Vydra provider requests so RFC2544 fake-IP proxy opt-ins reach generation calls. Fixes #79716.
**QQBot:** Route gateway WebSocket connections through the ambient proxy agent so deployments with `https_proxy`, `HTTPS_PROXY`, or `HTTP_PROXY` can reach the QQ gateway.
**Agents/subagents:** Treat `sessions_spawn` `model: "default"` as the default-model fallback and ignore ACP-only stream targets for native sub-agent spawns. Fixes #72078.
**Agents/failover:** Stop retrying assistant-prefill format rejections across auth profiles or model fallbacks, surfacing the deterministic provider error instead of requeueing the lane. Fixes #79688.
**Google/Gemini:** Resolve missing Gemini 3 Flash catalog rows through the Google provider template path so image-capable media-understanding models keep `input: ["text", "image"]` instead of falling back to text-only metadata. Fixes #79750. Thanks @fenglanhua and @hclsys.
**Memory/QMD:** Warn with a manual stale collection removal hint when QMD reports a path/pattern conflict but `collection list` lacks verifiable metadata, avoiding unsafe stderr-only rebinds. Refs #71783.
**🔒 Models/auth:** Make `openclaw models status --check` and dashboard auth health honor effective auth profile order while keeping stale profiles visible.
**Agents/failover:** Classify bare `stream_read_error` streaming failures as transient timeouts so configured model fallback runs instead of surfacing the raw transport error. Fixes #79689.
**Release/plugin publishing:** Retry transient ClawHub CLI dependency install failures, keep preview-passing plugins publishable when one preview cell flakes, and verify every expected ClawHub package version after publish so maintenance releases are faster to recover and less likely to hide partial plugin publishes.
**OpenAI:** Support `openai/chat-latest` as an explicit direct API-key model override for trying the moving ChatGPT Instant API alias without changing the stable default model.
**CLI:** Include computed `status` in `cron list --json` and `cron show --json` output so external tooling can read disabled/running/ok/error/skipped/idle state without reimplementing cron status derivation.
**CLI:** Make `openclaw channels list` channel-only, add `--all` for bundled and catalog channels, render installed/configured/enabled state, and move model auth/usage details to `openclaw models auth list`, `openclaw status`, and `openclaw models list`.
**CLI:** Honor owner enforcement for native command handlers.
**Active Memory:** Require admin scope for global memory toggles.
**Gateway/sessions:** Clear cached skills snapshots during `/new` and `sessions.reset` so long-lived channel sessions rebuild the visible skill list after skills change.
Auto-reply: gate inline skill tool dispatch through before-tool-call authorization hooks.
**Tavily:** Resolve dedicated `tavily_search` and `tavily_extract` tool credentials from the active runtime config snapshot, so `exec` SecretRef-backed API keys do not reach the tools unresolved.
**Plugins/install:** Use the same absolute POSIX npm lifecycle shell for managed plugin install, rollback, repair, and uninstall npm operations as staged package updates, preventing restricted PATH shells from breaking cleanup.
**Doctor/OpenAI Codex:** Revert the 2026.5.5 `doctor --fix` repair that rewrote valid `openai-codex/*` ChatGPT/Codex OAuth routes to `openai/*`, which could break OAuth-only GPT-5.5 setups or accidentally move users onto the OpenAI API-key route. If 2026.5.5 already changed your default model, run `openclaw models set openai-codex/gpt-5.5 && openclaw config validate` to switch the default agent back to the Codex OAuth PI route. Fixes #78407.
**Plugins/runtime fetch:** Drop third-party symbol metadata from plain request header dictionaries before passing them into native `fetch` or `Headers`, so SDK and guarded/proxy fetch paths do not reject otherwise valid plugin requests. Fixes #77846.
**Debug proxy:** Normalize captured fetch header dictionaries before replaying requests so symbol metadata from caller-owned header objects cannot make debug-proxy fetches fail.
**UI:** Bound guarded dispatcher cleanup after request timeouts so timed-out fetches return tool errors instead of leaving Gateway tool lanes active.
**Feishu:** Hydrate missing native topic starter thread IDs before session routing so first turns and follow-ups stay in the same topic session. Fixes #78262.
**LINE:** Reject `dmPolicy: "open"` configs without wildcard `allowFrom` so webhook DMs fail validation instead of being acknowledged and silently blocked before inbound processing. Fixes #78316.
**Telegram/Codex:** Keep message-tool-only progress drafts visible and render native Codex tool progress once per tool instead of duplicating item/tool draft lines. Fixes #75641.
**Providers/xAI:** Stop sending OpenAI-style reasoning effort controls to native Grok Responses models, so `xai/grok-4.3` no longer fails live Docker/Gateway runs with `Invalid reasoning effort`.
**Providers/xAI:** Clamp the bundled xAI thinking profile to `off` so live Gateway runs cannot send unsupported reasoning levels to native Grok Responses models.
**Matrix/approvals:** Retry approval delivery up to 3 times with a short backoff so transient Matrix send failures do not strand pending approval prompts.
**Discord/gateway:** Measure heartbeat ACK timeouts from the actual heartbeat send, preventing late initial heartbeats from triggering false reconnect loops while the channel is still awaiting readiness. Fixes #77668. Thanks @bryce-d-greybeard and @NikolaFC.
**Discord/guilds:** Route plain text control commands such as `/steer` through the normal authorization and mention gate instead of silently dropping them before an agent session can see them. Fixes #78080.
**UI:** Make the compaction count a compact `N Checkpoint(s)` disclosure and show expanded session-level details with modern checkpoint history cards across responsive table layouts.
**UI:** Keep chat and channel tabs responsive while history payloads and channel probes are slow, label partial channel status, and record slow chat/config render timings in the event log.
**Google Meet/Voice Call:** Make Twilio dial-in joins speak through the realtime Gemini voice bridge with paced audio streaming, backpressure-aware buffering, barge-in queue clearing, and no TwiML fallback during realtime speech, giving Meet participants a much snappier OpenClaw voice agent.
**Gateway/Windows:** Bind the default loopback gateway listener only to `127.0.0.1` on Windows so libuv's dual-stack `::1` behavior cannot wedge localhost HTTP requests. (#69701, fixes #69674).
**Plugins/migration:** Emit catalog-backed install hints when `plugins.entries` or `plugins.allow` references an official external plugin that is not installed, so upgraded configs point operators to `openclaw plugins install ` instead of telling them to remove valid plugin config.
**OpenAI/Codex media:** Advertise Codex audio transcription in runtime and manifest metadata and route active Codex chat models to the OpenAI transcription default instead of sending chat model ids to audio transcription.
**Dependencies:** Refresh runtime and provider packages including Pi 0.73.0, ACPX adapters, OpenAI, Anthropic, Slack, and TypeScript native preview, while keeping the Bedrock runtime installer override pinned below the Windows ARM Node 24 npm resolver failure.
**Agents/performance:** Pass the resolved workspace through BTW, compaction, embedded-run model generation, and PDF model setup so explicit agent-dir model refreshes can reuse the current workspace-scoped plugin metadata snapshot instead of falling back to cold plugin metadata scans.
**Plugins/performance:** Let unscoped model catalog and manifest-contract readers reuse the current workspace-compatible plugin metadata snapshot, avoiding repeated cold plugin metadata scans on hot control-plane paths while preserving env/config/workspace compatibility checks.
Config/plugin auto-enable: prefer the claiming plugin manifest id over a built-in channel alias when auto-allowlisting a configured channel, so WeCom/Yuanbao-style aliases **Fixed:** To the installed plugin id.
**Secrets/apply:** Preserve auth-profile `keyRef` and `tokenRef` fields when scrubbing provider-target secrets, so the canonical SecretRef metadata survives `secrets apply` without keeping plaintext values.
Plugins/active-memory: skip session-store channel entries that contain `:` when resolving the recall subagent's channel, so QQ c2c agent IDs (e.g. `c2c:10D4F7C2…`) and other scoped conversation IDs do not reach bundled-plugin `dirName` validation and crash the recall run. The same guard already applied to explicit `channelId` params; this extends it to store-derived channels.
...and 42 more
🐛 Bug Fixes
**Plugins/install:** Honor the beta update channel for onboarding and doctor-managed plugin installs by requesting floating npm and ClawHub specs with `@beta` while keeping persistent install records on the catalog default.
**WhatsApp/onboarding:** Canonicalize setup and pairing allowlist entries to WhatsApp's digit-only phone ids while still accepting E.164, JID, and `whatsapp:` inputs, so personal-phone allowlists match WhatsApp Web sender ids after setup.
**Gateway/startup:** Load provider plugins that own explicitly configured image, video, or music generation defaults so generation tools become live after gateway restart instead of remaining catalog-only. Fixes #77244. Thanks @buyuangtampan, @Nikoxx99, and @vincentkoc.
**Slack/subagents:** Keep resumed parent `message.send` calls in the originating Slack thread when ambient session thread context is present, and suppress successful silent child completion rows from follow-up findings.
**Slack/mentions:** Record thread participation for successful visible threaded Slack sends, including message-tool and media delivery paths, so unmentioned replies in bot-participated threads can bypass mention gating as documented. Fixes #77648.
**Infra/Windows:** Skip the POSIX `/tmp/openclaw` preferred path on Windows in `resolvePreferredOpenClawTmpDir` so log files, TTS temp files, and other writes land in `%TEMP%\openclaw-` instead of `C:\tmp\openclaw`. Fixes #60713.
**Media/Windows:** Open saved attachment temp files read/write before fsync so Windows WebChat and `chat.send` media offloads no longer fail with EPERM during durability flush.
**Agents/tools:** Honor narrow runtime tool allowlists when constructing embedded-runner tool families and bundled MCP/LSP runtimes, so cron/subagent runs that request tools such as `update_plan`, `browser`, `x_search`, channel login tools, or `group:plugins` no longer start with missing tools or unrelated bootstrap work.
**Codex plugin:** Mirror the experimental upstream app-server protocol and format generated TypeScript before drift checks, keeping OpenClaw's `experimentalApi` bridge compatible with latest Codex while preserving formatter gates.
**Telegram/media:** Derive no-caption inbound media placeholders from saved MIME metadata instead of the Telegram `photo` shape, so non-image and mixed attachments no longer reach the model as ``. Fixes #69793.
**Google Meet/Voice Call:** Make Twilio dial-in joins speak through the realtime Gemini voice bridge with paced audio streaming, backpressure-aware buffering, barge-in queue clearing, and no TwiML fallback during realtime speech, giving Meet participants a much snappier OpenClaw voice agent.
**Gateway/Windows:** Bind the default loopback gateway listener only to `127.0.0.1` on Windows so libuv's dual-stack `::1` behavior cannot wedge localhost HTTP requests. (#69701, fixes #69674).
**Plugins/migration:** Emit catalog-backed install hints when `plugins.entries` or `plugins.allow` references an official external plugin that is not installed, so upgraded configs point operators to `openclaw plugins install ` instead of telling them to remove valid plugin config.
**OpenAI/Codex media:** Advertise Codex audio transcription in runtime and manifest metadata and route active Codex chat models to the OpenAI transcription default instead of sending chat model ids to audio transcription.
**Dependencies:** Refresh runtime and provider packages including Pi 0.73.0, ACPX adapters, OpenAI, Anthropic, Slack, and TypeScript native preview, while keeping the Bedrock runtime installer override pinned below the Windows ARM Node 24 npm resolver failure.
**Agents/performance:** Pass the resolved workspace through BTW, compaction, embedded-run model generation, and PDF model setup so explicit agent-dir model refreshes can reuse the current workspace-scoped plugin metadata snapshot instead of falling back to cold plugin metadata scans.
**Plugins/performance:** Let unscoped model catalog and manifest-contract readers reuse the current workspace-compatible plugin metadata snapshot, avoiding repeated cold plugin metadata scans on hot control-plane paths while preserving env/config/workspace compatibility checks.
Config/plugin auto-enable: prefer the claiming plugin manifest id over a built-in channel alias when auto-allowlisting a configured channel, so WeCom/Yuanbao-style aliases **Fixed:** To the installed plugin id.
**Secrets/apply:** Preserve auth-profile `keyRef` and `tokenRef` fields when scrubbing provider-target secrets, so the canonical SecretRef metadata survives `secrets apply` without keeping plaintext values.
Plugins/active-memory: skip session-store channel entries that contain `:` when resolving the recall subagent's channel, so QQ c2c agent IDs (e.g. `c2c:10D4F7C2…`) and other scoped conversation IDs do not reach bundled-plugin `dirName` validation and crash the recall run. The same guard already applied to explicit `channelId` params; this extends it to store-derived channels.
...and 42 more
🐛 Bug Fixes
**Plugins/install:** Honor the beta update channel for onboarding and doctor-managed plugin installs by requesting floating npm and ClawHub specs with `@beta` while keeping persistent install records on the catalog default.
**WhatsApp/onboarding:** Canonicalize setup and pairing allowlist entries to WhatsApp's digit-only phone ids while still accepting E.164, JID, and `whatsapp:` inputs, so personal-phone allowlists match WhatsApp Web sender ids after setup.
**Gateway/startup:** Load provider plugins that own explicitly configured image, video, or music generation defaults so generation tools become live after gateway restart instead of remaining catalog-only. Fixes #77244. Thanks @buyuangtampan, @Nikoxx99, and @vincentkoc.
**Slack/subagents:** Keep resumed parent `message.send` calls in the originating Slack thread when ambient session thread context is present, and suppress successful silent child completion rows from follow-up findings.
**Slack/mentions:** Record thread participation for successful visible threaded Slack sends, including message-tool and media delivery paths, so unmentioned replies in bot-participated threads can bypass mention gating as documented. Fixes #77648.
**Infra/Windows:** Skip the POSIX `/tmp/openclaw` preferred path on Windows in `resolvePreferredOpenClawTmpDir` so log files, TTS temp files, and other writes land in `%TEMP%\openclaw-` instead of `C:\tmp\openclaw`. Fixes #60713.
**Media/Windows:** Open saved attachment temp files read/write before fsync so Windows WebChat and `chat.send` media offloads no longer fail with EPERM during durability flush.
**Agents/tools:** Honor narrow runtime tool allowlists when constructing embedded-runner tool families and bundled MCP/LSP runtimes, so cron/subagent runs that request tools such as `update_plan`, `browser`, `x_search`, channel login tools, or `group:plugins` no longer start with missing tools or unrelated bootstrap work.
**Codex plugin:** Mirror the experimental upstream app-server protocol and format generated TypeScript before drift checks, keeping OpenClaw's `experimentalApi` bridge compatible with latest Codex while preserving formatter gates.
**Telegram/media:** Derive no-caption inbound media placeholders from saved MIME metadata instead of the Telegram `photo` shape, so non-image and mixed attachments no longer reach the model as ``. Fixes #69793.
**Google Meet/Voice Call:** Make Twilio dial-in joins speak through the realtime Gemini voice bridge with paced audio streaming, backpressure-aware buffering, barge-in queue clearing, and no TwiML fallback during realtime speech, giving Meet participants a much snappier OpenClaw voice agent.
**Plugins/migration:** Emit catalog-backed install hints when `plugins.entries` or `plugins.allow` references an official external plugin that is not installed, so upgraded configs point operators to `openclaw plugins install ` instead of telling them to remove valid plugin config.
**OpenAI/Codex media:** Advertise Codex audio transcription in runtime and manifest metadata and route active Codex chat models to the OpenAI transcription default instead of sending chat model ids to audio transcription.
**Dependencies:** Refresh runtime and provider packages including Pi 0.73.0, ACPX adapters, OpenAI, Anthropic, Slack, and TypeScript native preview, while keeping the Bedrock runtime installer override pinned below the Windows ARM Node 24 npm resolver failure.
**Agents/performance:** Pass the resolved workspace through BTW, compaction, embedded-run model generation, and PDF model setup so explicit agent-dir model refreshes can reuse the current workspace-scoped plugin metadata snapshot instead of falling back to cold plugin metadata scans.
**Plugins/performance:** Let unscoped model catalog and manifest-contract readers reuse the current workspace-compatible plugin metadata snapshot, avoiding repeated cold plugin metadata scans on hot control-plane paths while preserving env/config/workspace compatibility checks.
Config/plugin auto-enable: prefer the claiming plugin manifest id over a built-in channel alias when auto-allowlisting a configured channel, so WeCom/Yuanbao-style aliases **Fixed:** To the installed plugin id.
**Secrets/apply:** Preserve auth-profile `keyRef` and `tokenRef` fields when scrubbing provider-target secrets, so the canonical SecretRef metadata survives `secrets apply` without keeping plaintext values.
Plugins/active-memory: skip session-store channel entries that contain `:` when resolving the recall subagent's channel, so QQ c2c agent IDs (e.g. `c2c:10D4F7C2…`) and other scoped conversation IDs do not reach bundled-plugin `dirName` validation and crash the recall run. The same guard already applied to explicit `channelId` params; this extends it to store-derived channels.
**Secrets/external channel contracts:** Also look in `/dist/` when resolving the `secret-contract-api` sidecar, so npm-published externalized channel plugins (e.g. `@openclaw/discord` since 2026.5.2) whose compiled artifacts live under `dist/` actually contribute their channel SecretRef contracts to the runtime snapshot. Without this, env-backed `channels.discord.token` SecretRefs silently failed to resolve at gateway start on 2026.5.3, leaving the channel `not configured` even though #76449 had landed the generic external-contract loader.
...and 40 more
🐛 Bug Fixes
**Infra/Windows:** Skip the POSIX `/tmp/openclaw` preferred path on Windows in `resolvePreferredOpenClawTmpDir` so log files, TTS temp files, and other writes land in `%TEMP%\openclaw-` instead of `C:\tmp\openclaw`. Fixes #60713.
**Media/Windows:** Open saved attachment temp files read/write before fsync so Windows WebChat and `chat.send` media offloads no longer fail with EPERM during durability flush.
**Agents/tools:** Honor narrow runtime tool allowlists when constructing embedded-runner tool families and bundled MCP/LSP runtimes, so cron/subagent runs that request tools such as `update_plan`, `browser`, `x_search`, channel login tools, or `group:plugins` no longer start with missing tools or unrelated bootstrap work.
**Codex plugin:** Mirror the experimental upstream app-server protocol and format generated TypeScript before drift checks, keeping OpenClaw's `experimentalApi` bridge compatible with latest Codex while preserving formatter gates.
**Telegram/media:** Derive no-caption inbound media placeholders from saved MIME metadata instead of the Telegram `photo` shape, so non-image and mixed attachments no longer reach the model as ``. Fixes #69793.
**Agents/cache:** Keep per-turn runtime context out of ordinary chat system prompts while still delivering hidden current-turn context, restoring prompt-cache reuse on chat continuations. Fixes #77431.
**Gateway/startup:** Include resolved thinking and fast-mode defaults in the `agent model` startup log line, defaulting unset startup thinking to `medium` without mixing in reasoning visibility.
**Agents/Tools:** Add post-compaction loop guard in `pi-embedded-runner` that arms after auto-compaction-retry and aborts the run with `compaction_loop_persisted` when the agent emits the same `(tool, args, result)` triple `windowSize` times (default 3) within that window. Disable via existing `tools.loopDetection.enabled`; tune via `tools.loopDetection.postCompactionGuard.windowSize`. Targets the failure mode where context-overflow + compaction does not break a tool-call loop. Refs #77474; carries forward #21597.
**Gateway/watch:** Suppress sync-I/O trace output during `pnpm gateway:watch --benchmark` unless explicitly requested, so CPU profiling no longer floods the terminal with stack traces.
**Gateway/watch:** When benchmark sync-I/O tracing is explicitly enabled, tee trace blocks to the benchmark output log and filter them from the terminal pane while keeping normal Gateway logs visible.
Plugins/file-transfer: Added **Bundled file-transfer plugin with `file_fetch`, `dir_list`, `dir_fetch`, and `file_write` agent**s for binary file ops on paired nodes; default-deny per-node path policy under `plugins.entries.file-transfer.config.nodes` with operator approval, symlink traversal refused by default (opt-in `followSymlinks`), and a 16 MB byte ceiling per round-trip.
**Google Meet/Voice Call:** Make Twilio dial-in joins speak through the realtime Gemini voice bridge with paced audio streaming, backpressure-aware buffering, barge-in queue clearing, and no TwiML fallback during realtime speech, giving Meet participants a much snappier OpenClaw voice agent.
**UI:** Show the active agent name in dashboard breadcrumbs without adding the current session key, keeping non-chat views oriented without crowding the topbar.
**UI:** Make the New Job sidebar collapsible so the jobs list can reclaim space while keeping the form one click away.
**Gateway/startup:** Keep model-catalog test helpers, run-session lookup code, QR pairing helpers, and TypeBox memory-tool schema construction out of hot startup import paths, reducing default gateway benchmark plugin-load and memory pressure.
**UI:** Record browser long animation frame or long task entries in the debug event log when supported, making slow dashboard renders easier to attribute from the UI.
**Channels/streaming:** Add unified `streaming.mode: "progress"` drafts with auto single-word status labels and shared progress configuration across Discord, Telegram, Matrix, Slack, and Microsoft Teams.
**Slack/streaming:** Add `streaming.progress.render: "rich"` for Block Kit progress drafts backed by structured progress line data.
**Slack/streaming:** Keep the newest rich progress lines when Block Kit limits trim long progress drafts.
**Channels/streaming:** Cap progress-draft tool lines by default so edited progress boxes avoid jumpy reflow from long wrapped lines.
**Plugins/discovery:** Ignore managed npm plugin packages that only expose TypeScript source entries without compiled runtime output, so stale/broken installs cannot hide a working bundled or reinstallable channel plugin during setup.
**CLI:** Treat OpenClaw stable correction versions like `2026.5.3-1` as newer than their base stable release, so package updates no longer ask for downgrade confirmation.
**Plugins/install:** Suppress dangerous-pattern scanner warnings for trusted official OpenClaw npm installs, so installing `@openclaw/discord` no longer prints credential-harvesting warnings for the official package.
**Plugins/release:** Make the published npm runtime verifier reject blank `openclaw.runtimeExtensions` entries instead of treating them as absent and passing via inferred outputs.
**🔒 Plugins/security:** Ignore inline and block comments when matching source-rule context in plugin install scans, so comment-only `fetch`/`post` references near environment defaults do not block clean plugins.
**Doctor/plugins:** Remove stale managed install records for bundled plugins even when the bundled plugin is not explicitly configured, so doctor cleanup cannot leave orphaned install metadata behind.
**UI:** Scope provider fallback cache entries by the selected fetch provider so config reloads cannot reuse another provider's cached fallback payload.
**UI:** Honor late-bound `tools.web.search.enabled: false` during tool execution so config reloads cannot leave an already-created `web_search` tool runnable.
Plugins/file-transfer: Added **Bundled file-transfer plugin with `file_fetch`, `dir_list`, `dir_fetch`, and `file_write` agent**s for binary file ops on paired nodes; default-deny per-node path policy under `plugins.entries.file-transfer.config.nodes` with operator approval, symlink traversal refused by default (opt-in `followSymlinks`), and a 16 MB byte ceiling per round-trip.
**Plugins/install:** Harden official plugin install, uninstall, update, onboarding, ClawHub fallback, npm dependency-state reporting, and beta-channel update paths so externalized plugins behave like first-class package installs.
**Gateway/performance:** Trim startup and Control UI hot paths by lazy-loading plugin/runtime discovery, cron, schema, shutdown, sessions, and model metadata work only when needed.
**Channels/replies:** Improve Discord status reactions and degraded transport reporting, add WhatsApp Channel/Newsletter targets, and tighten Telegram, Feishu, Matrix, Microsoft Teams, and Slack delivery/recovery behavior.
**Install/update:** Recover broken macOS LaunchAgent upgrades, reject source-only plugin packages before runtime load, and repair stale Gateway/plugin state during updates and doctor runs.
**Agent/runtime reliability:** Preserve streamed provider replies, delayed A2A session replies, prompt/tool delivery, memory recall, web search provider discovery, and provider-specific thinking/model metadata across common edge cases.
**Channels/streaming:** Add unified `streaming.mode: "progress"` drafts with auto single-word status labels and shared progress configuration across Discord, Telegram, Matrix, Slack, and Microsoft Teams.
**CLI:** Add `/steer ` for queue-independent steering of the active current-session run without starting a new turn when the session is idle.
**Tools/BTW:** Add `/side` as a text and native slash-command alias for `/btw` side questions.
**Doctor/config:** `doctor --fix` now commits safe legacy migrations even when unrelated validation issues (e.g. a missing plugin) prevent full validation from passing, so `agents.defaults.llm` and other known-legacy keys are always cleaned up by `doctor --fix` regardless of other config problems. Fixes #76798.
...and 18 more
🐛 Bug Fixes
**Channels/WhatsApp:** Allow `@whiskeysockets/libsignal-node` in `onlyBuiltDependencies` so pnpm v9+ `blockExoticSubdeps` no longer rejects the baileys git-tarball subdep and silences all inbound agent replies. Fixes #76539. Thanks @ottodeng and @vincentkoc.
**Gateway/systemd:** Preserve operator-added secrets in the Gateway env file across re-stage while clearing OpenClaw-managed keys (such as `OPENCLAW_GATEWAY_TOKEN`) so a fresh staging value is never shadowed by a stale env-file copy; operator secrets are also retained when the state-dir `.env` is empty. Fixes #76860.
**Plugin updates:** Do not short-circuit trusted official npm updates as unchanged when the default/latest spec still resolves to an already-installed prerelease that the installer should replace with a stable fallback.
**Plugin tools:** Keep auth-unavailable optional tools hidden even when another default tool from the same plugin is available and `tools.alsoAllow` names the optional tool.
**Realtime transcription:** Report socket closes before provider readiness as closed-before-ready failures instead of mislabeling them as connection timeouts for OpenAI, xAI, and Deepgram streaming transcription.
**OpenAI/Google Meet:** Fail realtime voice connection attempts when the socket closes before `session.updated`, avoiding stuck Meet joins waiting on a bridge that never became ready.
**QA/cache:** Require the full `CACHE-OK ` marker before live cache probes stop retrying, so suffix-only prose cannot hide a broken probe response.
**Slack/Matrix:** Avoid creating blank progress-draft messages when `streaming.progress.label=false` and progress tool lines are disabled.
**QA/Matrix:** Keep the mock OpenAI tool-progress provider aligned with exact-marker Matrix prompts so the hardened live preview scenario still forces a deterministic read before final delivery.
**OpenAI/Google Meet:** Wait for realtime voice `session.updated` before treating the bridge as connected, so Meet joins do not return with audio queued behind an unconfigured realtime session.
Plugins/file-transfer: Added **Bundled file-transfer plugin with `file_fetch`, `dir_list`, `dir_fetch`, and `file_write` agent**s for binary file ops on paired nodes; default-deny per-node path policy under `plugins.entries.file-transfer.config.nodes` with operator approval, symlink traversal refused by default (opt-in `followSymlinks`), and a 16 MB byte ceiling per round-trip.
**Plugins/install:** Harden official plugin install, uninstall, update, onboarding, ClawHub fallback, npm dependency-state reporting, and beta-channel update paths so externalized plugins behave like first-class package installs.
**Gateway/performance:** Trim startup and Control UI hot paths by lazy-loading plugin/runtime discovery, cron, schema, shutdown, sessions, and model metadata work only when needed.
**Channels/replies:** Improve Discord status reactions and degraded transport reporting, add WhatsApp Channel/Newsletter targets, and tighten Telegram, Feishu, Matrix, Microsoft Teams, and Slack delivery/recovery behavior.
**Install/update:** Recover broken macOS LaunchAgent upgrades, reject source-only plugin packages before runtime load, and repair stale Gateway/plugin state during updates and doctor runs.
**Agent/runtime reliability:** Preserve streamed provider replies, delayed A2A session replies, prompt/tool delivery, memory recall, web search provider discovery, and provider-specific thinking/model metadata across common edge cases.
**Channels/streaming:** Add unified `streaming.mode: "progress"` drafts with auto single-word status labels and shared progress configuration across Discord, Telegram, Matrix, Slack, and Microsoft Teams.
**CLI:** Add `/steer ` for queue-independent steering of the active current-session run without starting a new turn when the session is idle.
**Tools/BTW:** Add `/side` as a text and native slash-command alias for `/btw` side questions.
**Doctor/config:** `doctor --fix` now commits safe legacy migrations even when unrelated validation issues (e.g. a missing plugin) prevent full validation from passing, so `agents.defaults.llm` and other known-legacy keys are always cleaned up by `doctor --fix` regardless of other config problems. Fixes #76798.
...and 18 more
🐛 Bug Fixes
**Channels/WhatsApp:** Allow `@whiskeysockets/libsignal-node` in `onlyBuiltDependencies` so pnpm v9+ `blockExoticSubdeps` no longer rejects the baileys git-tarball subdep and silences all inbound agent replies. Fixes #76539. Thanks @ottodeng and @vincentkoc.
**Gateway/systemd:** Preserve operator-added secrets in the Gateway env file across re-stage while clearing OpenClaw-managed keys (such as `OPENCLAW_GATEWAY_TOKEN`) so a fresh staging value is never shadowed by a stale env-file copy; operator secrets are also retained when the state-dir `.env` is empty. Fixes #76860.
**Plugin updates:** Do not short-circuit trusted official npm updates as unchanged when the default/latest spec still resolves to an already-installed prerelease that the installer should replace with a stable fallback.
**Plugin tools:** Keep auth-unavailable optional tools hidden even when another default tool from the same plugin is available and `tools.alsoAllow` names the optional tool.
**Realtime transcription:** Report socket closes before provider readiness as closed-before-ready failures instead of mislabeling them as connection timeouts for OpenAI, xAI, and Deepgram streaming transcription.
**OpenAI/Google Meet:** Fail realtime voice connection attempts when the socket closes before `session.updated`, avoiding stuck Meet joins waiting on a bridge that never became ready.
**QA/cache:** Require the full `CACHE-OK ` marker before live cache probes stop retrying, so suffix-only prose cannot hide a broken probe response.
**Slack/Matrix:** Avoid creating blank progress-draft messages when `streaming.progress.label=false` and progress tool lines are disabled.
**QA/Matrix:** Keep the mock OpenAI tool-progress provider aligned with exact-marker Matrix prompts so the hardened live preview scenario still forces a deterministic read before final delivery.
**OpenAI/Google Meet:** Wait for realtime voice `session.updated` before treating the bridge as connected, so Meet joins do not return with audio queued behind an unconfigured realtime session.
Plugins/file-transfer: Added **Bundled file-transfer plugin with `file_fetch`, `dir_list`, `dir_fetch`, and `file_write` agent**s for binary file ops on paired nodes; default-deny per-node path policy under `plugins.entries.file-transfer.config.nodes` with operator approval, symlink traversal refused by default (opt-in `followSymlinks`), and a 16 MB byte ceiling per round-trip.
**Plugins/install:** Harden official plugin install, uninstall, update, onboarding, ClawHub fallback, npm dependency-state reporting, and beta-channel update paths so externalized plugins behave like first-class package installs.
**Gateway/performance:** Trim startup and Control UI hot paths by lazy-loading plugin/runtime discovery, cron, schema, shutdown, sessions, and model metadata work only when needed.
**Channels/replies:** Improve Discord status reactions and degraded transport reporting, add WhatsApp Channel/Newsletter targets, and tighten Telegram, Feishu, Matrix, Microsoft Teams, and Slack delivery/recovery behavior.
**Install/update:** Recover broken macOS LaunchAgent upgrades, reject source-only plugin packages before runtime load, and repair stale Gateway/plugin state during updates and doctor runs.
**Agent/runtime reliability:** Preserve streamed provider replies, delayed A2A session replies, prompt/tool delivery, memory recall, web search provider discovery, and provider-specific thinking/model metadata across common edge cases.
**Plugins/onboarding:** Let Manual setup install optional official plugins, including ClawHub-backed diagnostics with npm fallback, and expose the external Codex plugin as a selectable provider setup choice.
**CLI:** Include package dependency install state in `openclaw plugins list --json`, try plugin `@beta` updates first on the beta OpenClaw channel, and fall back to default/latest when no plugin beta release exists.
**Plugins/ClawHub:** Annotate 429 errors with reset windows and unauthenticated higher-rate-limit hints, so operators can tell when downloads recover and when signing in helps.
**Gateway/performance:** Lazy-load early runtime discovery, shutdown hooks, cron, channel-config schema metadata, restart sentinels, and maintenance timers after readiness; trim duplicate plugin auto-enable work and add startup CPU/profile controls.
...and 8 more
🐛 Bug Fixes
**Plugins/update:** Keep beta-installed OpenClaw package updates on the beta plugin channel even when config still says stable, so Discord and other externalized plugins update from compiled `@beta` packages instead of stale source-only `latest` artifacts.
**Agents/tools:** Stop treating `tools.deny: ["write"]` as an implicit `apply_patch` deny; operators who want to block patch writes should deny `apply_patch` or `group:fs` explicitly. Fixes #76749. Thanks @Nek-12 and @hclsys.
**Plugins/release:** Verify published plugin npm tarballs expose compiled runtime entries after publish, catching TS-only package artifacts before release closeout.
**CLI:** Exit cleanly with a nonzero status when message-command plugin registry loading fails before dispatch, preventing `openclaw-message` children from staying alive after plugin load errors. Fixes #76168.
**Gateway/update:** Recover an installed-but-unloaded macOS LaunchAgent after package updates, rerun Gateway health/version/channel readiness checks, and print restart, reinstall, and rollback guidance before reporting update failure.
**CLI:** Explain when a missing plugin command alias belongs to a bundled plugin that is disabled by default, including the `openclaw plugins enable ` repair command.
**Google Meet:** Route stateful CLI session commands through the gateway-owned runtime so joined realtime sessions survive after the starting CLI process exits. Fixes #76344.
**Memory/status:** Split builtin sqlite-vec store readiness from embedding-provider readiness in `memory status --deep` and `openclaw status`, so local vector-store failures no longer look like provider failures and provider failures no longer hide a healthy local vector store.
**CLI:** Trust a ready gateway memory probe when CLI-side active memory backend resolution is unavailable, preventing false "No active memory plugin is registered" warnings for healthy runtime setups. Fixes #76792.
**Memory/status:** Keep plain `openclaw memory status` and `openclaw memory status --json` on the cheap read-only path by reserving vector and embedding provider probes for `--deep` or `--index`. Fixes #76769.
External plugin installation, update, doctor repair, dependency reporting, and artifact metadata now cover the npm-first cutover, stale configured installs, missing package payloads, and beta-channel plugin fallback.
Gateway and agent hot paths are leaner across startup, session listing, task maintenance, prompt prep, plugin loading, tool descriptor planning, filesystem guards, and large runtime configs.
Control UI and WebChat are more resilient across Sessions, Cron, long-running Gateway WebSockets, grouped-message width, slash-command feedback, iOS PWA bounds, selection contrast, and Talk diagnostics.
Messaging fixes cover WhatsApp Channel/Newsletter targets, Telegram topic commands and networking, Discord delivery/startup edge cases, Slack threads, Signal groups/media, and visible reply routing.
Provider and media fixes cover OpenAI-compatible TTS/Realtime, OpenRouter/DeepSeek replay, Anthropic-compatible streaming, LM Studio reasoning metadata, Brave/SearXNG/Firecrawl web search, media paths, music, and voice-call routing.
**Gateway/startup and restart:** Skip plugin-backed auth-profile overlays during startup secrets preflight, reducing gateway readiness latency while keeping reload and OAuth recovery paths overlay-capable; add `openclaw gateway restart --force` and `--wait `, log active task run IDs before restart deferral timers, and report timeout restarts as explicit forced restarts.
**Plugins/ClawHub:** Make diagnostics, onboarding, doctor repair, and channel setup carry ClawPack metadata through install records while keeping explicit `clawhub:` installs on ClawHub and bare package installs on npm for the launch cutover.
**CLI:** Include package dependency install state in `openclaw plugins list --json` so scripts can spot missing plugin dependencies without runtime-loading plugins.
**Plugins/update:** On the beta OpenClaw update channel, default-line npm and ClawHub plugin updates try `@beta` first and fall back to default/latest when no plugin beta release exists.
**Plugins/runtime:** Scope broad runtime preloads to the effective plugin ids derived from config, startup planning, configured channels, slots, and auto-enable rules instead of importing every discoverable plugin.
...and 32 more
🐛 Bug Fixes
**Agents/OpenAI:** Default GPT-5 API-key sessions to the SSE Responses transport unless WebSocket is explicitly selected, restoring replies in fresh Control UI and WebChat beta installs where the auto WebSocket path connected but produced no model events.
**Agents/sessions:** Preserve terminal lifecycle state when final run metadata persists from a stale in-memory snapshot, preventing sessions from staying stuck as running after completed or timed-out turns.
**CLI:** Make `openclaw gateway start` repair stale managed service definitions that point at old OpenClaw versions, missing binaries, or temporary installer paths before starting; add concrete service, config, listener-owner, and log collection next steps when gateway probes fail and Bonjour finds no local gateway; avoid repeated plugin tool descriptor config hashing so large runtime configs do not block reply startup and trigger reconnect/timeouts. Refs #49012. Thanks @vincentkoc and @joshavant.
**Plugins/update/config:** Stop treating the non-plugin `auth` command root as a bundled plugin id, keep packaged upgrades and beta external plugin installs on stable runtime aliases and matching prerelease npm specs, detect tracked plugin install records whose package directories disappeared during `openclaw update`, reinstall them before normal plugin updates, fail the update if install records still point at missing disk payloads, and validate configured web-search providers plus statically suppressed model/provider pairs against the active plugin set at config load.
Codex/app-server: **Fixed:** Managed binaries from bundled `dist` chunks and from the `@openai/codex` package bin when installs do not provide a nearby `.bin/codex` shim, avoiding false missing-binary startup failures.
**Status:** Show the `openai-codex` OAuth profile for `openai/gpt-*` sessions running through the native Codex runtime instead of reporting auth as unknown.
**Status/update:** Resolve beta update-channel checks from the installed version when config still says `stable`, show configured channels in `openclaw status` and config-only `openclaw channels status` output even when the Gateway is unreachable, and let `status --deep` reuse live gateway channel credential state instead of warning on command-path-only token misses.
**Plugins/externalization:** Add official npm-first catalogs for externalized channel, provider, and generic plugins; install official external web-search plugins before saving provider config; repair missing configured, selected-search, and env-selected plugin installs from npm by default; keep official install docs, update examples, live Codex checks, diagnostics ClawHub packages, and persisted bundled-plugin relocation on default npm tags; keep Matrix and Mattermost bundled until their npm packages cut over; and keep ACPX, Google Chat, and LINE publishable plugin dist trees out of the core package while ClawHub pack files roll out.
**Plugins/ClawHub/source/registry:** Use the ClawHub artifact resolver response as the install decision before downloading, keep bare plugin package specs on npm for the launch cutover and reserve ClawHub resolution for explicit `clawhub:` specs until ClawHub pack readiness is deployed, discover source-only plugins such as Codex from `extensions/*`, install ClawPack artifacts from the explicit npm-pack `.tgz` resolver path, persist artifact kind, npm integrity, shasum, and tarball metadata for update/diagnostics flows, fall back to version metadata when the artifact resolver route is missing, keep the Docker ClawHub fixture aligned with npm-pack artifact resolution, explain unavailable explicit ClawHub ClawPack artifact downloads with a temporary npm install hint, and hash manifest/package metadata when validating persisted plugin registries so fast same-size rewrites cannot leave stale plugin metadata trusted.
**UI:** Add validated `gateway.controlUi.chatMessageMaxWidth` instead of patched bundled CSS, ignore malformed persisted cron rows before they enter UI state, guard stale cron render paths, and bound the default Sessions tab query to recent activity and fewer rows while keeping filters editable. Fixes #67935, #55047, #54439, and #76050; supersedes #54550 and #54552. Thanks @xiew4589-lang and @Neomail2.
External plugin installation now covers diagnostics, onboarding, doctor repair, channel setup, install/update records, and artifact metadata while keeping bare package installs on npm for the first cutover.
Gateway startup, session listing, task maintenance, prompt prep, plugin loading, and filesystem hot paths get targeted cache and fanout reductions for large or plugin-heavy installs.
Control UI and WebChat reliability improves across Sessions, Cron, long-running Gateway WebSockets, grouped-message width, slash-command feedback, iOS PWA bounds, selection contrast, and Talk diagnostics.
Channel and provider fixes cover Telegram topic commands and networking, Discord delivery and startup edge cases, OpenAI-compatible TTS/Realtime, OpenRouter/DeepSeek replay, Anthropic-compatible streaming, Brave/SearXNG/Firecrawl web search, and voice-call routing.
**Gateway/startup:** Skip plugin-backed auth-profile overlays during startup secrets preflight, reducing gateway readiness latency while keeping reload and OAuth recovery paths overlay-capable.
**Plugins/ClawHub:** Make diagnostics, onboarding, doctor repair, and channel setup carry ClawPack metadata through install records while keeping explicit `clawhub:` installs on ClawHub and bare package installs on npm for the launch cutover.
**CLI:** Include package dependency install state in `openclaw plugins list --json` so scripts can spot missing plugin dependencies without runtime-loading plugins.
**Plugins/runtime:** Scope broad runtime preloads to the effective plugin ids derived from config, startup planning, configured channels, slots, and auto-enable rules instead of importing every discoverable plugin.
**Agents/runtime:** Reuse the startup-loaded plugin registry for request-time providers, tools, channel actions, web/capability/memory/migration helpers, and memoized provider extra-params so stable embedded-run inputs no longer repeat plugin registry resolution while model-specific transport hook patches stay isolated.
**Agents/runtime:** Memoize transcript replay-policy resolution for stable config and process-env runs while preserving custom-env provider hook behavior.
...and 43 more
🐛 Bug Fixes
**Agents/OpenAI:** Default GPT-5 API-key sessions to the SSE Responses transport unless WebSocket is explicitly selected, restoring replies in fresh Control UI and WebChat beta installs where the auto WebSocket path connected but produced no model events.
**Agents/sessions:** Preserve terminal lifecycle state when final run metadata persists from a stale in-memory snapshot, preventing sessions from staying stuck as running after completed or timed-out turns.
**CLI:** Make `openclaw gateway start` repair stale managed service definitions that point at old OpenClaw versions, missing binaries, or temporary installer paths before starting.
**Updates/plugins:** Keep packaged upgrades and beta external plugin installs on stable runtime aliases and matching prerelease npm specs, avoiding stale WebChat runtime chunks and old Twitch packages after upgrading from 2026.4.29.
Codex/app-server: **Fixed:** Managed binaries from bundled `dist` chunks and from the `@openai/codex` package bin when installs do not provide a nearby `.bin/codex` shim, avoiding false missing-binary startup failures.
**Status:** Show the `openai-codex` OAuth profile for `openai/gpt-*` sessions running through the native Codex runtime instead of reporting auth as unknown.
**Plugins/ClawHub:** Use the ClawHub artifact resolver response as the install decision before downloading, keeping legacy ZIP fallback and future ClawPack npm-pack installs on the same explicit resolver path.
**Plugins/ClawHub:** Keep bare plugin package specs on npm for the launch cutover and reserve ClawHub resolution for explicit `clawhub:` specs until ClawHub pack readiness is deployed.
**Plugins/source checkout:** Discover source-only plugins such as Codex from the `extensions/*` workspace while using npm package excludes as the packaged-core boundary, removing the stale core-bundle metadata path.
**Plugins/ClawHub:** Install ClawPack artifacts from the explicit npm-pack `.tgz` resolver path and persist artifact kind, npm integrity, shasum, and tarball metadata for update and diagnostics flows.
External plugin installation now covers diagnostics, onboarding, doctor repair, channel setup, install/update records, and artifact metadata while keeping bare package installs on npm for the first cutover.
Gateway startup, session listing, task maintenance, prompt prep, plugin loading, and filesystem hot paths get targeted cache and fanout reductions for large or plugin-heavy installs.
Control UI and WebChat reliability improves across Sessions, Cron, long-running Gateway WebSockets, grouped-message width, slash-command feedback, iOS PWA bounds, selection contrast, and Talk diagnostics.
Channel and provider fixes cover Telegram topic commands and networking, Discord delivery and startup edge cases, OpenAI-compatible TTS/Realtime, OpenRouter/DeepSeek replay, Anthropic-compatible streaming, Brave/SearXNG/Firecrawl web search, and voice-call routing.
**Gateway/startup:** Skip plugin-backed auth-profile overlays during startup secrets preflight, reducing gateway readiness latency while keeping reload and OAuth recovery paths overlay-capable.
**Plugins/ClawHub:** Make diagnostics, onboarding, doctor repair, and channel setup carry ClawPack metadata through install records while keeping explicit `clawhub:` installs on ClawHub and bare package installs on npm for the launch cutover.
**CLI:** Include package dependency install state in `openclaw plugins list --json` so scripts can spot missing plugin dependencies without runtime-loading plugins.
**Plugins/runtime:** Scope broad runtime preloads to the effective plugin ids derived from config, startup planning, configured channels, slots, and auto-enable rules instead of importing every discoverable plugin.
**Agents/runtime:** Reuse the startup-loaded plugin registry for request-time providers, tools, channel actions, web/capability/memory/migration helpers, and memoized provider extra-params so stable embedded-run inputs no longer repeat plugin registry resolution while model-specific transport hook patches stay isolated.
**Agents/runtime:** Memoize transcript replay-policy resolution for stable config and process-env runs while preserving custom-env provider hook behavior.
...and 43 more
🐛 Bug Fixes
**Agents/OpenAI:** Default GPT-5 API-key sessions to the SSE Responses transport unless WebSocket is explicitly selected, restoring replies in fresh Control UI and WebChat beta installs where the auto WebSocket path connected but produced no model events.
**Agents/sessions:** Preserve terminal lifecycle state when final run metadata persists from a stale in-memory snapshot, preventing sessions from staying stuck as running after completed or timed-out turns.
**CLI:** Make `openclaw gateway start` repair stale managed service definitions that point at old OpenClaw versions, missing binaries, or temporary installer paths before starting.
**Updates/plugins:** Keep packaged upgrades and beta external plugin installs on stable runtime aliases and matching prerelease npm specs, avoiding stale WebChat runtime chunks and old Twitch packages after upgrading from 2026.4.29.
Codex/app-server: **Fixed:** Managed binaries from bundled `dist` chunks and from the `@openai/codex` package bin when installs do not provide a nearby `.bin/codex` shim, avoiding false missing-binary startup failures.
**Status:** Show the `openai-codex` OAuth profile for `openai/gpt-*` sessions running through the native Codex runtime instead of reporting auth as unknown.
**Plugins/ClawHub:** Use the ClawHub artifact resolver response as the install decision before downloading, keeping legacy ZIP fallback and future ClawPack npm-pack installs on the same explicit resolver path.
**Plugins/ClawHub:** Keep bare plugin package specs on npm for the launch cutover and reserve ClawHub resolution for explicit `clawhub:` specs until ClawHub pack readiness is deployed.
**Plugins/source checkout:** Discover source-only plugins such as Codex from the `extensions/*` workspace while using npm package excludes as the packaged-core boundary, removing the stale core-bundle metadata path.
**Plugins/ClawHub:** Install ClawPack artifacts from the explicit npm-pack `.tgz` resolver path and persist artifact kind, npm integrity, shasum, and tarball metadata for update and diagnostics flows.
Messaging and automation get active-run steering by default, visible-reply enforcement, spawned subagent routing metadata, and opt-in follow-up commitments for heartbeat-delivered reminders. Thanks @vincentkoc, @scoootscooob, @samzong, and @vignesh07.
Memory grows into a people-aware wiki with provenance views, per-conversation Active Memory filters, partial recall on timeout, and bounded REM preview diagnostics. Thanks @vincentkoc, @quengh, @joeykrug, and @samzong.
Provider/model coverage expands with NVIDIA onboarding/catalogs plus **Performance:** Faster manifest-backed model/auth paths, Bedrock Opus 4.7 thinking parity, and safer Codex/OpenAI-compatible replay and streaming behavior. Thanks @eleqtrizit, @shakkernerd, @prasad-yashdeep, @woodhouse-bot, and @LyHug.
Gateway and packaged-plugin reliability focuses on slow-host startup, reusable model catalogs, event-loop readiness diagnostics, runtime-dependency repair, stale-session recovery, and version-scoped update caches. Thanks @lpendeavors, @DerFlash, @vincentkoc, @pashpashpash, and @jhsmith409.
Channel fixes cluster around Slack Block Kit limits, Telegram proxy/webhook/polling/send resilience, Discord startup/rate-limit handling, WhatsApp delivery/liveness, and Microsoft Teams/Matrix/Feishu edge cases. Thanks @slackapi, @SymbolStar, @djgeorg3, @TinyTb, @dseravalli, @nklock, and @alex-xuweilong.
Security and operations add OpenGrep scanning, sharper GHSA triage policy, safer exec/pairing/owner-scope handling, Docker/onboarding automation, and web-fetch IPv6 ULA opt-in for trusted proxy stacks. Thanks @jesse-merhi, @pgondhi987, @mmaps, @jinjimz, and @jeffrey701.
**🔒 Security/tools:** Configured tool sections (`tools.exec`, `tools.fs`) no longer implicitly widen restrictive profiles (`messaging`, `minimal`). Users who need those tools under a restricted profile must add explicit `alsoAllow` entries; a startup warning identifies affected configs. Fixes #47487.
**Agents/commitments:** Add opt-in inferred follow-up commitments with hidden batched extraction, per-agent/per-channel scoping, heartbeat delivery, CLI management, a simple `commitments.enabled`/`commitments.maxPerDay` config, and heartbeat-interval due-time clamping so magical check-ins do not echo immediately.
**Messages/queue:** Make `steer` drain all pending Pi steering messages at the next model boundary, keep legacy one-at-a-time steering as `queue`, and add a dedicated steering queue docs page.
**Messages/queue:** Default active-run queueing to `steer` with a 500ms followup fallback debounce, and document the queue modes, precedence, and drop policies on the command queue page.
...and 20 more
🐛 Bug Fixes
**Providers/OpenAI Codex:** Preserve existing wrapped Codex streams during OpenAI attribution so PI OAuth bearer injection reaches ChatGPT/Codex Responses, and strip native Codex-only unsupported payload fields without touching custom compatible endpoints.
Agents/tool-result guard: use the resolved runtime context token budget for non-context-engine tool-result overflow checks, so long tool-heavy sessions no longer compact early when `contextTokens` is larger than native `contextWindow`. Fixes #74917.
**Gateway/systemd:** Exit with sysexits 78 for supervised lock and `EADDRINUSE` conflicts so `RestartPreventExitStatus=78` stops `Restart=always` restart loops instead of repeatedly reloading plugins against an occupied port. Fixes #75115.
**Agents/runtime:** Skip blank visible user prompts at the embedded-runner boundary before provider submission while still allowing internal runtime-only turns and media-only prompts, so Telegram/group sessions no longer leak raw empty-input provider errors when replay history exists. Fixes #74137. Thanks @yelog, @Gracker, and @nhaener.
Auto-reply/group chats: fall back to automatic source delivery when a channel precomputes message-tool-only replies but the `message` tool is unavailable, so Discord/Slack-style group turns do not silently complete without a visible reply. Fixes #74868.
**Browser/gateway:** Share one browser control runtime across the HTTP control server and `browser.request`, and refresh browser profile config from the source snapshot, so CLI status/start honors configured `browser.executablePath`, `headless`, and `noSandbox` instead of falling back to stale auto-detection. Fixes #75087; repairs #73617. Thanks @civiltox and @martingarramon.
**Agents/subagents:** Bound automatic orphan recovery with persisted recovery attempts and a wedged-session tombstone, and teach task maintenance/doctor to reconcile those sessions so restart loops no longer require manual `sessions.json` surgery. Fixes #74864.
**Gateway/startup:** Skip pre-bind web-fetch provider discovery for credential-free `tools.web.fetch` config, so Docker/Kubernetes gateways bind even when optional fetch limits are present. Fixes #74896.
**Infra/tmp:** Tolerate concurrent temp-dir permission repairs by rechecking directories that another process already tightened, so parallel ACP subprocess startup no longer throws `Unsafe fallback OpenClaw temp dir`. Fixes #66867. Thanks @Kane808-AI and @jarvisz8.
**Signal:** Match group allowlists against inbound Signal group ids as well as sender ids, and process explicitly configured Signal groups without requiring mentions unless `requireMention` is set. Fixes #53308. Thanks @minupla and @juan-flores077.