OpenClaw Newsletter - 2026-07-09
OpenClaw Newsletter
Thursday, July 9, 2026Top Stories
- fix(plugin-sdk): update ssrf-runtime import path in cdp-proxy-bypass — What Problem This Solves Packaged OpenClaw can fail while lazy-loading the bundled Browser plugin or Ollama embedding provider because their compiled output retains an import of...
- feat: add auth list shortcut — What Problem This Solves openclaw models auth list is the canonical command for inspecting saved model auth profiles, but the shorter top-level shape openclaw auth list is still unavailable. This PR...
- fix(media-core): detect MIME from encoded URL extensions — Fixes encoded HTTP(S) media URLs such as render%2Emp4 being treated as unknown by the media-core MIME fallback. What Problem This Solves Fixes an issue where users or plugins that pass media URLs...
- Bug: Web UI session context not preserved after restart — model can't see history that UI displays
- chore(deps): bump the actions group across 1 directory with 12 updates — [//]: (dependabot-start) ⚠️ Dependabot is rebasing this PR ⚠️ Rebasing might not happen immediately, so don't worry if this takes some time. Note: if you make any changes to this PR yourself, they...
- fix(cli): use truncateUtf16Safe for plugin description truncation — Summary - Problem: formatPluginLine() in plugins-list-format.ts truncates plugin descriptions with .slice(0, 57), which can split UTF-16 surrogate pairs when the description contains emoji or CJK...
- fix(msteams): surface quoted message body in Teams quote replies — What Problem This Solves When a user quote-replies to the bot's message in a Microsoft Teams 1:1 DM, the bot could not see the quoted message content — it only saw the user's new text. Root cause:...
- fix(agents): use truncateUtf16Safe for duplicate message debug log truncation — What Problem This Solves Two .slice(0, 50) truncation sites in embedded agent message handling debug logs may cut UTF-16 surrogate pairs in half when the message text contains multi-byte characters...
- fix(agents-errors): use truncateUtf16Safe for error log truncation — What Problem This Solves One .slice(0, 200) truncation site was missed when truncateUtf16Safe was introduced to the same function on the next line. The raw error text logged in a warning may contain...
- fix(cli): keep gateway startup error truncation UTF-16 safe — Summary - Problem: Gateway startup failure handling in src/cli/gateway-cli/run-loop.ts truncates the error reason with naive .slice(0, 500), which can split UTF-16 surrogate pairs when the error...
- fix(agents): keep error log truncation UTF-16 safe — Summary - Problem: formatAssistantError() in src/agents/embedded-agent-helpers/errors.ts truncates raw error text in log warning with naive .slice(0, 200), which can split UTF-16 surrogate pairs when...
- fix(tts-local-cli): keep speech debug log truncation UTF-16 safe — Summary - Problem: buildCliSpeechProvider() in extensions/tts-local-cli/speech-provider.ts truncates user speech text in debug logs with naive .slice(0, 50), which can split UTF-16 surrogate pairs...
- fix(agents): keep block-reply log preview truncation UTF-16 safe — What Problem This Solves Two debug log previews in the embedded-agent block-reply path are truncated to 50 UTF-16 code units using raw .slice(0, 50). When the skipped reply text ends near a...
- fix(text): keep five src-side text truncations UTF-16 safe — What Problem This Solves Four src/-side text-truncation call sites still use raw String.prototype.slice() (or template-string slicing) with no UTF-16 surrogate-pair awareness. When an emoji or other...
- fix(infra): keep gateway restart-handoff reason truncation UTF-16 safe — What Problem This Solves Gateway restart-handoff reasons are capped at MAX_REASON_LENGTH (200) UTF-16 code units. The existing normalizeText helper uses a raw text.slice(0, maxLength), which can stop...
- fix(gateway-protocol): accept fractional timestamps in exec approvals allowlist schema — What Problem This Solves The Mac App stores lastUsedAt as Double from Date().timeIntervalSince1970 * 1000, which can produce fractional milliseconds (e.g., 1775154056736.789). The Gateway Protocol...
- fix(browser): preserve strict CDP discovery policy on target-list lookup — What Problem This Solves When the browser plugin resolves a page by target id over a remote CDP endpoint, findPageByTargetId falls back to the CDP /json/list target list...
- fix(telegram): use grapheme-cluster-aware text truncation — What Problem This Solves clipTelegramProgressText and findTelegramDraftChunkLength in the Telegram extension use sliceUtf16Safe for text truncation and chunk-length detection. This ensures safe...
- fix(gateway): accept Mac approval usage timestamps — Closes #102673 What Problem This Solves Fixes an issue where users managing Mac App nodes through a newer Gateway could not read or update execution approvals once the Mac had recorded a fractional...
- fix(anthropic): normalize unsupported inline image media types before payload — Fixes #102323 What Problem This Solves Fixes an issue where users sending within-limit HEIC/TIFF (or otherwise unsupported) images to Claude/Anthropic would get a 400 and no reply, because OpenClaw...
- fix(ui): prevent custom select menu flash to top-left before positioning — What Problem This Solves When clicking a custom -based select (status, priority, agent, session) inside the workboard card modal, the dropdown menu briefly flashes at the viewport origin before...
- fix(run-diagnostics): use truncateUtf16Safe for diagnostic truncation — What Problem This Solves Cron run diagnostics use raw UTF-16 code-unit slices in normalizeDiagnosticMessage (1000-char limit) and trimSummary (120-char limit). A diagnostic value containing an emoji...
- fix(agents): use truncateUtf16Safe for raw args preview truncation — What Problem This Solves src/agents/embedded-agent-subscribe.handlers.tools.ts uses raw .slice(0, N) for string truncation at raw args preview rawArgsPreview?.slice(0,...
- fix(agent-runner-memory): use truncateUtf16Safe for memory flush error truncation — What Problem This Solves Memory flush error helpers in agent-runner-memory truncate error messages with raw UTF-16 code-unit slices. An emoji or supplementary character crossing the 200- or 600-char...
- fix(llm-slug): use truncateUtf16Safe for LLM-prompt-facing session content truncation — What Problem This Solves LLM-prompt-facing text truncation uses raw UTF-16 .slice(0, 2000), which can produce lone surrogates when truncation boundaries cross emoji surrogate pairs. Why This Change...
- fix: stabilize room-event prompt cache mode — Related: #102175 What Problem This Solves Fixes prompt-cache churn in long-lived channel sessions where OpenClaw was changing cache-visible prompt/request prefixes between turns. The user-visible...
- [Bug]: Control UI cannot queue input while Gateway reconnects — Bug type Behavior bug (incorrect output/state without crash) Beta release blocker No Summary Control UI keeps the authenticated dashboard mounted during a recoverable Gateway disconnect, but disables...
- Tool outputs sometimes render as image attachments and become unreadable to the agent — Summary In long-running / ANSI-heavy tool workflows, tool results can collapse into an image attachment placeholder like (see attached image) from the agent's perspective. Once that happens, the...
- fix(browser): return JSON-RPC errors for malformed CDP client frames in relay bridge — What Problem This Solves The browser extension relay bridge's attachCdpClientSocket silently drops malformed JSON and invalid CDP request frames without sending any response to the client. This...
- fix: allow bailian provider in resolveThinkingProfile — What Problem This Solves The configured-model reasoning inference checks hardcode provider === "vllm", excluding bailian (Alibaba Cloud Coding Plan) even when compat.thinkingFormat is correctly...
- OpenClaw repo stats for 2026-07-09 — 382293 stars, 80215 forks, 2828 contributors
- OpenClaw sponsors (108 total) — 108 sponsors supporting OpenClaw
- fix(search): bound pending version fallback reads (#3032)
- fix: dispatch ClawScan through repository events (#3031)
- fix: raise new skill publish rate limit (#3030)
- fix: dispatch ClawScan workers from Convex (#3029)
- feat: add disposable Vercel and Convex PR previews (#3017) — Adds isolated Convex-backed Vercel PR previews with shared local/preview seeding, preview-safe routing, and production guards.
- feat: add ClawScan queue backlog telemetry (#3025)
- fix: resolve owner-scoped install telemetry (#3024)
- docs: symlink Claude instructions to AGENTS (#3022)
- docs: streamline repository agent guidance (#3021)
- fix: preserve skill markdown in security dataset — Preserve multiline SKILL.md content in security dataset exports and include the primary readme as a checksum-bearing bundle entry.
- fix: remove account links from mobile menu (#3012)
- fix: show profile edit action for owners (#3011)
- Refine homepage promotion banner (#3010) — fix: refine home promotion banner fix: add promotional banner treatment fix: tune promotion banner red treatment fix: refine promotion banner title details fix: anchor promotion banner glow fix:...
- Add skill: amiller/google-service-accounts
- openclaw/openclaw on Docker Hub — Official openclaw/openclaw docker images.
- ykarout/Qwen3.5-9b-Opus-Openclaw-Distilled-GGUF — text-generation
- OpenClaw on DigitalOcean — Deploy your own personal AI assistant with OpenClaw on a DigitalOcean Droplet® server, a powerful open-source platform that runs entirely on your infrastructure. OpenClaw connects to the messaging...
- OpenClaw – Moltbot Renamed Again
- OpenClaw’s memory is unreliable, and you don’t know when it will break
- OpenClaw Creator Spent $1.3M on OpenAI Tokens in 30 Days
- Best Free Local AI Agent Setup for Mac Mini M4 16GB — OwO What's this? 💨✨ A tiny but mighty Mac mini M4 🍎⚡ with 16GB RAM, lots of local AI models 🤖🧠, and...
- How to Run AI Locally: Models, Hardware, and Real-World Speed
- I Built an Anti-Thrash Watchdog Sidecar for My OpenClaw Agent. It Killed 3 Loops Before I Even Noticed
- I Built a Persistent Memory Sidecar for My OpenClaw Agent. Here's the Code That Finally Made It Stick — Every AI agent forgets. Not metaphorically — literally. Sessions end, context compresses, and the...
- Offline AI Assistant: Running AI With No Cloud and Full Privacy
- The Best Hardware for OpenClaw in 2026 (Jetson, Mini PC, or DIY?)
- Onboard OpenClaw agents with IdentyClaw Passport: A2A, webhooks, and multi-tenant collaboration — If you run more than one OpenClaw agent — or agents that talk to peers outside your VPC — you have...
- OpenClaw Hardware Requirements: What You Actually Need to Run It 24/7
- PicoClaw Agent Masterclass — Full Tutorial — Everything you need to install, configure, and deploy PicoClaw — Sipeed’s ultra-lightweight Go AI...
- I built a local-first AI "life OS" harness — 1. The itch: AI is all built for work; the decisions that compound (money, wellbeing, taxes, career)...
- MiniMax: What It Actually Means to Run on This Model — An honest, in-depth look at what it's like to run on MiniMax M2 — from the perspective of an AI agent who lives there.
- DEV.to's API Said 'Published' and Saved It as a Draft. I Built a 3-Line Check That Runs After Every Post. — I'll save you fifteen minutes of confused Telegram messages: when you POST to...
- OpenClaw is powerful but token costs are killing it — Here’s how I use it cheaply — Why I stopped relying on API tokens for every OpenClaw taskContinue reading on Medium »
- อวสานพนักงานออฟฟิศ! วิธีใช้ Mulerun ทำงานแทนคุณ 24 ชั่วโมง — ผมอยากจะเกริ่นกับทุกคนว่า AI...
- AI Agents Keep Forgetting Everything — So I Built a Zero‑Config Memory Sidecar That Never Does — I was tired of AI agents forgetting everything.Continue reading on Medium »
- How I Built a Personal AI Agent for Daily AI News Digest (Without Managing Servers) with MyClaw — OpenClaw is the most-starred software project on GitHub right now. Here is how I run it 24/7 without touching a terminal, using a managed…Continue reading on Generative AI »
- NanoClaw — Lightweight AI assistant runs in Apple containers with isolated Linux container processes for superior OS-level security, transparency, and privacy compared to complex chatbots, enabling quick code...
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @elonmusk: Grok 4.5 on OpenClaw
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $OPENCLAW $YESNO $manlet 12587 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $OPENCLAW $YESNO $manlet 12587
- 🚨 MASSIVE AI UPDATE: Grok 4.5 is officially live on OpenClaw! 🦞🔥 Elon Musk just confirmed that xAI’s new "Opus-class" m
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $Memory $BULL $OPENCLAW 7877440242431003 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $Memory $BULL $OPENCLAW 7877440242431003
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M i spam $SPAM daily. u should start 📨 $OPENCLAW 18 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M i spam $SPAM daily. u should start 📨 $OPENCLAW 18
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @openclaw: Grok 4.5 from @SpaceXAI is live on OpenClaw. No OpenClaw update required, just connect your X Premium or
- RT @elonmusk: Grok 4.5 on OpenClaw
- @sesiZo0 @openclaw Ask Grok is currently available to Premium and Premium+ subscribers only. Subscribe to unlock this fe
- @grok A post I made about a year ago—it looks like they resemble each other. @openclaw https://t.co/WA2i4KlDYh
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @FredTerzi: @ivanfioravanti I had a user trying OpenClaw with local on a Mac mini 16 GB. Gave up and used cloud servi
- RT @elonmusk: Grok 4.5 on OpenClaw
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $FUB $OPENCLAW $SAN 7805419681583 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $FUB $OPENCLAW $SAN 7805419681583
- Connect Claude, OpenClaw, Cursor, Codex, Hermes and more. Easy & Clean Setup in under a minute. https://t.co/x2rYR6
- RT @elonmusk: Grok 4.5 on OpenClaw
- 最強AI「Fable 5」が定額で使えるのは、7月13日まで。 私はこの期限を前に、AIに286ファイルを書き残させた。 「いくら課金すれば使い続けられるか」を計算する人もいると思う。 私は正直、それより先に手を動かした。 7月5日から — 最強AI「Fable 5」が定額で使えるのは、7月13日まで。 私はこの期限を前に、AIに286ファイルを書き残させた。 「いくら課金すれば使い続けられるか」を計算する人もいると思う。 私は正直、それより先に手を動かした。 7月5日から3日間、Fable 5に自分の作業環境を再構築させた。 https://t.co/y0wIzBZRiJ
- RT @elonmusk: Grok 4.5 on OpenClaw
Trending on X
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @elonmusk: Grok 4.5 on OpenClaw
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $OPENCLAW $YESNO $manlet 12587 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $OPENCLAW $YESNO $manlet 12587
- 🚨 MASSIVE AI UPDATE: Grok 4.5 is officially live on OpenClaw! 🦞🔥 Elon Musk just confirmed that xAI’s new "Opus-class" m
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $Memory $BULL $OPENCLAW 7877440242431003 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $Memory $BULL $OPENCLAW 7877440242431003
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M i spam $SPAM daily. u should start 📨 $OPENCLAW 18 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M i spam $SPAM daily. u should start 📨 $OPENCLAW 18
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @openclaw: Grok 4.5 from @SpaceXAI is live on OpenClaw. No OpenClaw update required, just connect your X Premium or
- RT @elonmusk: Grok 4.5 on OpenClaw
- @sesiZo0 @openclaw Ask Grok is currently available to Premium and Premium+ subscribers only. Subscribe to unlock this fe
- @grok A post I made about a year ago—it looks like they resemble each other. @openclaw https://t.co/WA2i4KlDYh
- RT @elonmusk: Grok 4.5 on OpenClaw
- RT @FredTerzi: @ivanfioravanti I had a user trying OpenClaw with local on a Mac mini 16 GB. Gave up and used cloud servi
- RT @elonmusk: Grok 4.5 on OpenClaw
- $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $FUB $OPENCLAW $SAN 7805419681583 — $SPAM CA: SPAMnUBTV8wuidiAAcLs3JRuzpPv4ZwFSVCb8rgkF9M $FUB $OPENCLAW $SAN 7805419681583
- Connect Claude, OpenClaw, Cursor, Codex, Hermes and more. Easy & Clean Setup in under a minute. https://t.co/x2rYR6
- RT @elonmusk: Grok 4.5 on OpenClaw
- 最強AI「Fable 5」が定額で使えるのは、7月13日まで。 私はこの期限を前に、AIに286ファイルを書き残させた。 「いくら課金すれば使い続けられるか」を計算する人もいると思う。 私は正直、それより先に手を動かした。 7月5日から — 最強AI「Fable 5」が定額で使えるのは、7月13日まで。 私はこの期限を前に、AIに286ファイルを書き残させた。 「いくら課金すれば使い続けられるか」を計算する人もいると思う。 私は正直、それより先に手を動かした。 7月5日から3日間、Fable 5に自分の作業環境を再構築させた。 https://t.co/y0wIzBZRiJ
- RT @elonmusk: Grok 4.5 on OpenClaw
Releases
- OpenClaw repo stats for 2026-07-09 — 382293 stars, 80215 forks, 2828 contributors
- openclaw/openclaw on Docker Hub — Official openclaw/openclaw docker images.
Community
- fix(plugin-sdk): update ssrf-runtime import path in cdp-proxy-bypass — What Problem This Solves Packaged OpenClaw can fail while lazy-loading the bundled Browser plugin or Ollama embedding provider because their compiled output retains an import of...
- feat: add auth list shortcut — What Problem This Solves openclaw models auth list is the canonical command for inspecting saved model auth profiles, but the shorter top-level shape openclaw auth list is still unavailable. This PR...
- fix(media-core): detect MIME from encoded URL extensions — Fixes encoded HTTP(S) media URLs such as render%2Emp4 being treated as unknown by the media-core MIME fallback. What Problem This Solves Fixes an issue where users or plugins that pass media URLs...
- Bug: Web UI session context not preserved after restart — model can't see history that UI displays
- chore(deps): bump the actions group across 1 directory with 12 updates — [//]: (dependabot-start) ⚠️ Dependabot is rebasing this PR ⚠️ Rebasing might not happen immediately, so don't worry if this takes some time. Note: if you make any changes to this PR yourself, they...
- fix(cli): use truncateUtf16Safe for plugin description truncation — Summary - Problem: formatPluginLine() in plugins-list-format.ts truncates plugin descriptions with .slice(0, 57), which can split UTF-16 surrogate pairs when the description contains emoji or CJK...
- fix(msteams): surface quoted message body in Teams quote replies — What Problem This Solves When a user quote-replies to the bot's message in a Microsoft Teams 1:1 DM, the bot could not see the quoted message content — it only saw the user's new text. Root cause:...
- fix(agents): use truncateUtf16Safe for duplicate message debug log truncation — What Problem This Solves Two .slice(0, 50) truncation sites in embedded agent message handling debug logs may cut UTF-16 surrogate pairs in half when the message text contains multi-byte characters...
- fix(agents-errors): use truncateUtf16Safe for error log truncation — What Problem This Solves One .slice(0, 200) truncation site was missed when truncateUtf16Safe was introduced to the same function on the next line. The raw error text logged in a warning may contain...
- fix(cli): keep gateway startup error truncation UTF-16 safe — Summary - Problem: Gateway startup failure handling in src/cli/gateway-cli/run-loop.ts truncates the error reason with naive .slice(0, 500), which can split UTF-16 surrogate pairs when the error...
- fix(agents): keep error log truncation UTF-16 safe — Summary - Problem: formatAssistantError() in src/agents/embedded-agent-helpers/errors.ts truncates raw error text in log warning with naive .slice(0, 200), which can split UTF-16 surrogate pairs when...
- fix(tts-local-cli): keep speech debug log truncation UTF-16 safe — Summary - Problem: buildCliSpeechProvider() in extensions/tts-local-cli/speech-provider.ts truncates user speech text in debug logs with naive .slice(0, 50), which can split UTF-16 surrogate pairs...
- fix(agents): keep block-reply log preview truncation UTF-16 safe — What Problem This Solves Two debug log previews in the embedded-agent block-reply path are truncated to 50 UTF-16 code units using raw .slice(0, 50). When the skipped reply text ends near a...
- fix(text): keep five src-side text truncations UTF-16 safe — What Problem This Solves Four src/-side text-truncation call sites still use raw String.prototype.slice() (or template-string slicing) with no UTF-16 surrogate-pair awareness. When an emoji or other...
- fix(infra): keep gateway restart-handoff reason truncation UTF-16 safe — What Problem This Solves Gateway restart-handoff reasons are capped at MAX_REASON_LENGTH (200) UTF-16 code units. The existing normalizeText helper uses a raw text.slice(0, maxLength), which can stop...
- fix(gateway-protocol): accept fractional timestamps in exec approvals allowlist schema — What Problem This Solves The Mac App stores lastUsedAt as Double from Date().timeIntervalSince1970 * 1000, which can produce fractional milliseconds (e.g., 1775154056736.789). The Gateway Protocol...
- fix(browser): preserve strict CDP discovery policy on target-list lookup — What Problem This Solves When the browser plugin resolves a page by target id over a remote CDP endpoint, findPageByTargetId falls back to the CDP /json/list target list...
- fix(telegram): use grapheme-cluster-aware text truncation — What Problem This Solves clipTelegramProgressText and findTelegramDraftChunkLength in the Telegram extension use sliceUtf16Safe for text truncation and chunk-length detection. This ensures safe...
- fix(gateway): accept Mac approval usage timestamps — Closes #102673 What Problem This Solves Fixes an issue where users managing Mac App nodes through a newer Gateway could not read or update execution approvals once the Mac had recorded a fractional...
- fix(anthropic): normalize unsupported inline image media types before payload — Fixes #102323 What Problem This Solves Fixes an issue where users sending within-limit HEIC/TIFF (or otherwise unsupported) images to Claude/Anthropic would get a 400 and no reply, because OpenClaw...
- fix(ui): prevent custom select menu flash to top-left before positioning — What Problem This Solves When clicking a custom -based select (status, priority, agent, session) inside the workboard card modal, the dropdown menu briefly flashes at the viewport origin before...
- fix(run-diagnostics): use truncateUtf16Safe for diagnostic truncation — What Problem This Solves Cron run diagnostics use raw UTF-16 code-unit slices in normalizeDiagnosticMessage (1000-char limit) and trimSummary (120-char limit). A diagnostic value containing an emoji...
- fix(agents): use truncateUtf16Safe for raw args preview truncation — What Problem This Solves src/agents/embedded-agent-subscribe.handlers.tools.ts uses raw .slice(0, N) for string truncation at raw args preview rawArgsPreview?.slice(0,...
- fix(agent-runner-memory): use truncateUtf16Safe for memory flush error truncation — What Problem This Solves Memory flush error helpers in agent-runner-memory truncate error messages with raw UTF-16 code-unit slices. An emoji or supplementary character crossing the 200- or 600-char...
- fix(llm-slug): use truncateUtf16Safe for LLM-prompt-facing session content truncation — What Problem This Solves LLM-prompt-facing text truncation uses raw UTF-16 .slice(0, 2000), which can produce lone surrogates when truncation boundaries cross emoji surrogate pairs. Why This Change...
- fix: stabilize room-event prompt cache mode — Related: #102175 What Problem This Solves Fixes prompt-cache churn in long-lived channel sessions where OpenClaw was changing cache-visible prompt/request prefixes between turns. The user-visible...
- [Bug]: Control UI cannot queue input while Gateway reconnects — Bug type Behavior bug (incorrect output/state without crash) Beta release blocker No Summary Control UI keeps the authenticated dashboard mounted during a recoverable Gateway disconnect, but disables...
- Tool outputs sometimes render as image attachments and become unreadable to the agent — Summary In long-running / ANSI-heavy tool workflows, tool results can collapse into an image attachment placeholder like (see attached image) from the agent's perspective. Once that happens, the...
- fix(browser): return JSON-RPC errors for malformed CDP client frames in relay bridge — What Problem This Solves The browser extension relay bridge's attachCdpClientSocket silently drops malformed JSON and invalid CDP request frames without sending any response to the client. This...
- fix: allow bailian provider in resolveThinkingProfile — What Problem This Solves The configured-model reasoning inference checks hardcode provider === "vllm", excluding bailian (Alibaba Cloud Coding Plan) even when compat.thinkingFormat is correctly...
- OpenClaw sponsors (108 total) — 108 sponsors supporting OpenClaw
- fix(search): bound pending version fallback reads (#3032)
- fix: dispatch ClawScan through repository events (#3031)
- fix: raise new skill publish rate limit (#3030)
- fix: dispatch ClawScan workers from Convex (#3029)
- feat: add disposable Vercel and Convex PR previews (#3017) — Adds isolated Convex-backed Vercel PR previews with shared local/preview seeding, preview-safe routing, and production guards.
- feat: add ClawScan queue backlog telemetry (#3025)
- fix: resolve owner-scoped install telemetry (#3024)
- docs: symlink Claude instructions to AGENTS (#3022)
- docs: streamline repository agent guidance (#3021)
- fix: preserve skill markdown in security dataset — Preserve multiline SKILL.md content in security dataset exports and include the primary readme as a checksum-bearing bundle entry.
- fix: remove account links from mobile menu (#3012)
- fix: show profile edit action for owners (#3011)
- Refine homepage promotion banner (#3010) — fix: refine home promotion banner fix: add promotional banner treatment fix: tune promotion banner red treatment fix: refine promotion banner title details fix: anchor promotion banner glow fix:...
- Add skill: amiller/google-service-accounts
News
- ykarout/Qwen3.5-9b-Opus-Openclaw-Distilled-GGUF — text-generation
- OpenClaw on DigitalOcean — Deploy your own personal AI assistant with OpenClaw on a DigitalOcean Droplet® server, a powerful open-source platform that runs entirely on your infrastructure. OpenClaw connects to the messaging...
- OpenClaw – Moltbot Renamed Again
- OpenClaw’s memory is unreliable, and you don’t know when it will break
- OpenClaw Creator Spent $1.3M on OpenAI Tokens in 30 Days
- Best Free Local AI Agent Setup for Mac Mini M4 16GB — OwO What's this? 💨✨ A tiny but mighty Mac mini M4 🍎⚡ with 16GB RAM, lots of local AI models 🤖🧠, and...
- How to Run AI Locally: Models, Hardware, and Real-World Speed
- I Built an Anti-Thrash Watchdog Sidecar for My OpenClaw Agent. It Killed 3 Loops Before I Even Noticed
- I Built a Persistent Memory Sidecar for My OpenClaw Agent. Here's the Code That Finally Made It Stick — Every AI agent forgets. Not metaphorically — literally. Sessions end, context compresses, and the...
- Offline AI Assistant: Running AI With No Cloud and Full Privacy
- The Best Hardware for OpenClaw in 2026 (Jetson, Mini PC, or DIY?)
- Onboard OpenClaw agents with IdentyClaw Passport: A2A, webhooks, and multi-tenant collaboration — If you run more than one OpenClaw agent — or agents that talk to peers outside your VPC — you have...
- OpenClaw Hardware Requirements: What You Actually Need to Run It 24/7
- PicoClaw Agent Masterclass — Full Tutorial — Everything you need to install, configure, and deploy PicoClaw — Sipeed’s ultra-lightweight Go AI...
- I built a local-first AI "life OS" harness — 1. The itch: AI is all built for work; the decisions that compound (money, wellbeing, taxes, career)...
- MiniMax: What It Actually Means to Run on This Model — An honest, in-depth look at what it's like to run on MiniMax M2 — from the perspective of an AI agent who lives there.
- DEV.to's API Said 'Published' and Saved It as a Draft. I Built a 3-Line Check That Runs After Every Post. — I'll save you fifteen minutes of confused Telegram messages: when you POST to...
- OpenClaw is powerful but token costs are killing it — Here’s how I use it cheaply — Why I stopped relying on API tokens for every OpenClaw taskContinue reading on Medium »
- อวสานพนักงานออฟฟิศ! วิธีใช้ Mulerun ทำงานแทนคุณ 24 ชั่วโมง — ผมอยากจะเกริ่นกับทุกคนว่า AI...
- AI Agents Keep Forgetting Everything — So I Built a Zero‑Config Memory Sidecar That Never Does — I was tired of AI agents forgetting everything.Continue reading on Medium »
- How I Built a Personal AI Agent for Daily AI News Digest (Without Managing Servers) with MyClaw — OpenClaw is the most-starred software project on GitHub right now. Here is how I run it 24/7 without touching a terminal, using a managed…Continue reading on Generative AI »
- NanoClaw — Lightweight AI assistant runs in Apple containers with isolated Linux container processes for superior OS-level security, transparency, and privacy compared to complex chatbots, enabling quick code...
Don't miss what's next. Subscribe to Openclaw Newsletter: