OpenClaw Newsletter - 2026-02-06

In This Issue

Editorial Intro
Release Updates
New Skills Spotlight
Community Spotlight
Ecosystem Watch
In The Press
Security Corner

Editorial Intro

openclaw 2026.2.6 — Changes - Models: support Anthropic Opus 4.6 and OpenAI Codex gpt-5.3-codex (forward-compat fallbacks). (#9853, #10720, #9995) Thanks @TinyTb, @calvin-hpnet, @tyler6204. - Providers: add xAI (Grok)...
openclaw 2026.2.3 — Changes - Telegram: remove last @ts-nocheck from bot-handlers.ts, use Grammy types directly, deduplicate StickerMetadata. Zero @ts-nocheck remaining in src/telegram/. (#9206) - Telegram: remove...
openclaw 2026.2.2 — Changes - Feishu: add Feishu/Lark plugin support + docs. (#7313) Thanks @jiulingyun (openclaw-cn). - Web UI: add Agents dashboard for managing agent files, tools, skills, models, channels, and cron...
openclaw 2026.2.1 — Changes - Docs: onboarding/install/i18n/exec-approvals/Control UI/exe.dev/cacheRetention updates + misc nav/typos. (#3050, #3461, #4064, #4675, #4729, #4763, #5003, #5402, #5446, #5474, #5663, #5689,...
openclaw 2026.1.30 — Changes - CLI: add completion command (Zsh/Bash/PowerShell/Fish) and auto-setup during postinstall/onboarding. - CLI: add per-agent models status (--agent filter). (#4780) Thanks @jlowin. - Agents:...
openclaw 2026.1.29 — Status: stable. Changes - Rebrand: rename the npm package/CLI to openclaw, add a openclaw compatibility shim, and move extensions to the @openclaw/ scope. - Onboarding: strengthen security warning...
Clawdbot 2026.1.24 — 2026.1.24 Highlights - Providers: Ollama discovery + docs; Venice guide upgrades + cross-links. (#1606) Thanks @abhaymundhara. https://docs.clawd.bot/providers/ollama...
Clawdbot 2026.1.23 — 2026.1.23 Highlights - TTS: move Telegram TTS into core + enable model-driven TTS tags by default for expressive audio replies. (#1559) Thanks @Glucksberg. https://docs.clawd.bot/tts - Gateway: add...
clawdbot 2026.1.22 — Changes - Highlight: Compaction safeguard now uses adaptive chunking, progressive fallback, and UI status + retries. (#1466) Thanks @dlauer. - Providers: add Antigravity usage tracking to status...
clawdbot 2026.1.21 — Highlights - Lobster optional plugin tool for typed workflows + approval gates. docs - Custom assistant identity + avatars in the Control UI. docs docs - Cache optimizations: cache-ttl pruning +...
clawdbot 2026.1.20 — Changes - Control UI: add copy-as-markdown with error feedback. (#1345) https://docs.clawd.bot/web/control-ui - Control UI: drop the legacy list view. (#1345) https://docs.clawd.bot/web/control-ui -...
clawdbot 2026.1.16-2 — Highlights - Hooks: add hooks system with bundled hooks, CLI tooling, and docs. (#1028) — thanks @ThomsenDrake. https://docs.clawd.bot/hooks - Media: add inbound media understanding...
clawdbot 2026.1.15 — Highlights - Plugins: add provider auth registry + clawdbot models auth login for plugin-driven OAuth/API key flows. - Browser: improve remote CDP/Browserless support (auth passthrough, wss upgrade,...
clawdbot 2026.1.14-1 — Install - npm i -g clawdbot@2026.1.14-1 Highlights - Web search: web_search/web_fetch tools (Brave API) + first-time setup in onboarding/configure. - Browser control: Chrome extension relay takeover...
clawdbot 2026.1.12-2 — Fixes - Packaging: include dist/memory/** in the npm tarball (fixes ERR_MODULE_NOT_FOUND for dist/memory/index.js).
clawdbot 2026.1.12 — Highlights - BREAKING: rename chat “providers” (Slack/Telegram/WhatsApp/…) to channels across CLI/RPC/config; legacy config keys auto-migrate on load (and are written back as channels.). - Memory:...
clawdbot 2026.1.11-3 — Fixes - CLI: avoid top-level await warnings in the entrypoint on fresh installs. - CLI: show a commit hash in the banner for npm installs (package.json gitHead fallback).
clawdbot 2026.1.11-2 — Fixes - Installer: ensure the CLI entrypoint is executable after npm installs. - Packaging: include dist/plugins/ in the npm package to avoid missing module errors.
clawdbot 2026.1.11-1 — Fixes - Installer: include patches/ in the npm package so postinstall patching works for npm/bun installs.
clawdbot 2026.1.11 — Highlights - Plugins are now first-class: loader + CLI management, plus the new Voice Call plugin. - Config: modular $include support for split config files. (#731) — thanks @pasogott. - Agents/Pi:...
clawdbot 2026.1.10 — Highlights - CLI: clawdbot status now table-based + shows OS/update/gateway/daemon/agents/sessions; status --all adds a full read-only debug report (tables, log tails, Tailscale summary, and scan...
clawdbot 2026.1.9 — Highlights - Microsoft Teams provider: polling, attachments, outbound CLI send, per-channel policy. - Models/Auth expansion: OpenCode Zen + MiniMax API onboarding; token auth profiles + auth order;...
clawdbot 2026.1.8 — 2026.1.8 Highlights - Security: DMs locked down by default across providers; pairing-first + allowlist guidance. - Sandbox: per-agent scope defaults + workspace access controls; tool/session...
v2026.1.5-3 — Fixes - NPM package: include missing runtime dist folders (slack/signal/imessage/tui/wizard/control-ui/daemon) to avoid ERR_MODULE_NOT_FOUND in Node 25 npx installs.
clawdbot 2026.1.5-1 — Fixes - NPM package: include dist/sessions so clawdbot agent resolves session helpers in npx installs. - Node 25: avoid unsupported directory import by targeting...
clawdbot 2026.1.5 — Highlights - Models: add image-specific model config (agent.imageModel + fallbacks) and scan support. - Agent tools: new image tool routed to the image model (when configured). - Config: default...
clawdis 2.0.0-beta5 — Fixed - Media: preserve GIF animation when uploading to Discord/other providers (skip JPEG optimization for image/gif). - Agent runtime: update pi-mono dependencies to 0.31.1 (agent-core split). -...
clawdis 2.0.0-beta4 — 2.0.0-beta4 — 2025-12-27 Fixes - Package contents: include Discord/hooks build outputs in the npm tarball to avoid missing module errors.
clawdis 2.0.0-beta3 — Highlights - First-class Clawdis tools (browser, canvas, nodes, cron) replace the old clawdis- skills; tool schemas are now injected directly into the agent runtime. - Per-session model selection +...
clawdis 2.0.0-beta2 — Highlights - Bundled gateway packaging: bun-compiled embedded gateway, new gateway-daemon command, launchd support, DMG packaging (zip+DMG). - Skills platform: managed/bundled skills, install...
Cron jobs not firing on schedule — Cron jobs not firing on schedule Environment - OpenClaw 2026.2.3-1 (d84eb46) - macOS, gateway running since 9:16am PST Issue Multiple cron jobs failed to fire on Feb 6, 2026: 1. Daily Email Triage...
fix(session-status): pass session-level overrides to buildStatusMessage (#10867) — Summary Fixes #10867 — session_status tool was not passing session-level thinking/verbose/reasoning/elevated overrides to buildStatusMessage, causing it to fall back to agent defaults (e.g. "Think:...
[Bug]: Cron jobs with 'every' schedule never execute - recomputeNextRuns() pushes times forward — Summary Cron jobs with schedule.kind: "every" never execute automatically because recomputeNextRuns() always recalculates nextRunAtMs based on current time, pushing all scheduled times forward before...
Feature Request: Pre-compaction hook for agents to persist context — Problem When context compaction occurs, agents lose access to recent conversation history without warning. The compaction happens mid-session, and by the time memoryFlush runs, important context may...
docs: add model-level usage summary via CodexBar CLI — Summary Adds a small utility script and documentation support for summarizing per-model usage costs using CodexBar’s local cost logs. This helps users understand: - which models are used most -...
Browser restart bug: Old Chrome processes accumulate causing CDP timeouts — Issue Description OpenClaw browser control service has a restart bug that causes Chrome processes to accumulate, leading to resource conflicts and CDP communication timeouts. Environment - OS: macOS...
fix(media): resolve PDF being misidentified as text/plain in group chats — - PDF files with highly printable ASCII content (headers, fonts) were incorrectly detected as text/plain by looksLikeUtf8Text() - This caused extractFileContentFromSource() to skip PDF extraction,...
fix(discord): smarter retry for application ID resolution — Fixes #10887 Summary - Distinguish transient vs permanent failures in fetchDiscordApplicationId — return undefined for permanent errors (401, 403, bad token) and throw for transient ones (5xx, 429,...
fix: add missing clientTools parameter to runEmbeddedAttempt call — Summary Passes the clientTools parameter through to runEmbeddedAttempt, enabling client-defined tools in the gateway's OpenResponses /v1/responses endpoint. Repro Steps 1. Send a request to...
google-antigravity/claude-opus-4-6-thinking not discovered from Antigravity API — Description claude-opus-4-6-thinking is available in the Google Antigravity API but OpenClaw 2026.2.6 doesn't recognize it as a valid model. Steps to Reproduce 1. Authenticate with Google Antigravity...
ci: optimize CI pipeline — macOS consolidation, Windows perf, concurrency controls — Summary PR CI runs are severely bottlenecked by two issues: (1) macOS runner starvation — 4 macOS jobs per PR exhaust GitHub's org-wide 5-slot concurrency limit, leaving PRs queued for 23+ hours, and...
feat: wire message_sending hook in outbound delivery — Summary Connects the existing message_sending plugin hook to the outbound delivery path, allowing plugins to modify outgoing message content before it is sent to channels. Motivation The...
Thinking Level Inconsistencies in session_status Display — Description: When attempting to set the thinking level to 'medium' using the /reasoning medium command, the session_status command incorrectly reports the thinking level as "Think: off". The agent's...
Add .caf to AUDIO_FILE_EXTENSIONS for iMessage voice message transcription — Problem iMessage voice messages are stored as .caf (Core Audio Format) files. OpenClaw's audio transcription pipeline doesn't recognize them because .caf is not in the hardcoded AUDIO_FILE_EXTENSIONS...
[Bug]: WhatsApp channel provider not starting despite being enabled — WhatsApp channel provider not starting despite being enabled" Details: Gateway starts Telegram but skips WhatsApp entirely No startup attempt or error in logs Plugin is loaded, config is valid,...
fix: prevent thinking content leak when enforceFinalTag is active — Problem When a provider has isReasoningTagProvider=true (e.g. providers using tag enforcement), the model's "thinking out loud" content — internal reasoning, tool call traces, historical context...
fix(media): strip audio attachments after successful transcription — Summary Strips raw audio attachments from the model context when transcription has already succeeded. This saves 500-8000+ tokens per voice message that were previously wasted on binary data the...
[Feature]: Middleware Hook for Cost-Saving Model Routing (Local LLM Integration) — Summary High API costs associated with using top-tier models (like Claude 3.5 Sonnet or Opus) for every interaction. Currently, high-volume usage can exceed $60/day because routine tasks (reminders,...
[Bug]: no output — Summary What went wrong? Steps to reproduce 1. after the installation (skip skill and hooks setting), also endup with no output Expected behavior model respons Actual behavior What actually happened?...
[Feature]: Support .agents/skills directory for cross-agent skill discovery — Summary Support the emerging .agents/skills/ directory convention for per-project skill discovery, as proposed in agentskills/agentskills#15. Motivation Currently each AI coding tool uses its own...
feat: Support routingHint and improve URL routing for Anthropic (Claude Max) subscriptions — Summary This PR adds support for the routingHint parameter in Anthropic (Claude Max/Pro) subscription-based "token" authentication. It also improves the onboarding flow to optionally prompt for this...
iMessage voice messages (.caf) not recognized as audio — missing from AUDIO_FILE_EXTENSIONS — Bug iMessage voice messages use .caf (Core Audio Format), but .caf is not in the AUDIO_FILE_EXTENSIONS set in pi-embedded-helpers. The media pipeline never classifies these as audio, so transcription...
Refactor frame.sh for improved argument handling — Hi, I’ve added a small, self-contained Bash script that simplifies extracting a single frame from a video using ffmpeg. It includes basic validation and a clean CLI interface. Happy to adjust the...
CLI: add plugins uninstall command (#5985) — Add unistall plugin command closing https://github.com/openclaw/openclaw/issues/5985 This PR adds a new plugins uninstall CLI subcommand and implements the underlying uninstall logic in...
Docs: Add model-level usage summary via CodexBar CLI — What Add a beginner-friendly guide and script for summarizing per-model usage (Codex / Claude) using CodexBar's local cost logs. This includes: - current (most recent) model usage - full per-model...
Bug: Feishu channel text commands (/reset, /status) not recognized — Summary Text commands like /reset, /status, /new sent via Feishu channel are not recognized as commands. Instead, they are passed to the agent as regular messages. Environment - OpenClaw Version:...
feat(onboard): add Ollama provider support — Summary bash ◇ Onboarding mode │ QuickStart │ ◇ QuickStart ─────────────────────────╮ │ │ │ Gateway port: 18789 │ │ Gateway bind: Loopback (127.0.0.1) │ │ Gateway auth: Token (default) │ │ Tailscale...
feat(discord): support available_tags in channel-edit for forum tag management — Summary Add support for editing forum channel tags (available_tags) via the channel-edit message action. This enables programmatic management of forum tags — creating, renaming, reordering, and...
fix(skills): recursive directory filtering to actually exclude venv/node_modules — Summary Fix the skills directory filtering implementation from PR #10204 to actually work. Problem The original implementation symlinked entire skill directories: This meant venv/, node_modules/,...
feat(browser): session-aware context isolation for multi-agent browser use — Summary Add browser session isolation so multiple sessions/sub-agents can use the browser simultaneously without tab/state conflicts, while sharing login state (cookies) from the main context....
OpenClaw repo stats for 2026-02-06 — 172027 stars, 27832 forks, 444 contributors
fix: restore soft-deleted users on reauth (#106) (thanks @mkrokosz)
fix: Convert userId to string for targetId comparison — The auditLogs.targetId field is v.string() in the schema, so explicitly convert the Id to string to ensure type-safe comparison. Co-Authored-By: Claude Opus 4.5
fix: Restore soft-deleted users on re-authentication (with ban check) — Users who deleted their account were unable to sign in again with the same GitHub account. The OAuth flow would complete but the user would remain logged out due to the deletedAt field being set....
fix: update footer branding to OpenClaw (#122) (thanks @jontsai)
fix: update footer branding from Clawdbot to OpenClaw — - Change 'A Clawdbot project' to 'An OpenClaw project' - Update link from clawd.bot to openclaw.ai
fix: backfill empty handles in ensure (#158) (thanks @adlai88)
fix: handle empty-string handle in ensure() fallback — ?? (nullish coalescing) treats "" as a valid value, so users with handle="" never get a fallback derived from name/email. Change to || so empty strings fall through to the next candidate. This...
fix: self-healing VT scan queue — - Batch size 100 (was 10) - process more per run - Skip skills checked in last 60 min - avoid hammering same hashes - After 10 failed checks, mark as pending.scan.stale - drops from queue - Track...
fix: randomize VT scan queue + add health monitoring — - Fetch 5x batch size and shuffle to avoid queue head-blocking - Add getScanQueueHealthInternal to monitor queue status - Log warnings when queue is unhealthy (>50 pending or >24h stale) - Return...
fix: use yellow styling for suspicious code insight blocks — - Malicious verdicts keep red border/background - Suspicious verdicts now use yellow/amber to match badge color
feat: add VT rescan trigger and backfill function — - Add requestRescan() to trigger Code Insight via /analyse endpoint - Update pollPendingScans to request rescan when no Code Insight - Add backfillPendingScans for one-time backlog clearing
feat: add cron job to poll VT for pending scan results — - Add pollPendingScans action to vt.ts that checks VT for Code Insight verdicts - Add getPendingScanSkillsInternal query to get skills awaiting scan - Add vt-pending-scans cron job running every 5...
feat: VT Code Insight visibility matrix for malicious/suspicious skills — - Malicious skills: visible for transparency, downloads blocked via moderationFlags - Suspicious skills: visible with warning banner, downloads allowed - Neither appears in search/listings (not...
fix: show verdict labels (Benign/Suspicious/Malicious) instead of engine stats
Update description: Goodreads for OpenClaw agents
Add skill: andrewleonardi/shellf
Add gousto skill
Follow contributing guide.
Update install instructions
Add 913 skills
openclaw v2026.2.3-1 on npm — WhatsApp gateway CLI (Baileys web) with Pi RPC agent
openclaw-cli 2026.2.3-1 on Homebrew — Your own personal AI assistant
Tommy6688/openclaw
gguf-org/openclaw
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 is what Apple intelligence should have been
A sane but bull case on Clawdbot / OpenClaw
Nanobot: Ultra-Lightweight Alternative to OpenClaw
Ask HN: Any real OpenClaw (Clawd Bot/Molt Bot) users? What's your experience?
OpenClaw is basically a cascade of LLMs in prime position to mess stuff up
OpenClaw: When AI Agents Get Full System Access. Security nightmare?
OpenClaw security assessment [pdf]
Please stop using OpenClaw, formerly known as Moltbot
OpenClaw is everywhere all at once, and a disaster waiting to happen
Openclaw on Oracle's Free Tier: Always-On AI for $0/Month
Ask HN: OpenClaw users, what is your token spend?
SecureClaw – 100% free social-driven security platform for OpenClaw
Pi: The Minimal Agent Within OpenClaw
Show HN: Cheapest Managed OpenClaw hosting – 30s setup (agent37.com)
Ask HN: What serious task have you accomplished with Moltbot / OpenClaw?
From magic to malware: How OpenClaw's agent skills become an attack surface
Using Microvm.nix to Sandbox OpenClaw
Properly securing OpenClaw with authentication
Show HN: Prompt-injection firewall for OpenClaw agents
Heartbeats in OpenClaw: Cheap Checks First, Models Only When You Need Them — A practical OpenClaw pattern: use rule-based heartbeats most of the time, and only call an LLM (e.g. Haiku) when there’s real signal to summarise.
Add Persistent Memory to Claude Code with Mem0 (5-Minute Setup) — Claude Code is a phenomenal piece of technology. But it is affected by the same problem every LLM is...
Getting Started with OpenClaw (Clawdbot/Moltbot): A Developer's Guide to AI Assistants — 🦞 OpenClaw: The "2026" Field Guide to Your AI Sidekick A practical walkthrough of...
January 2026 DigitalOcean Tutorial Roundup: OpenClaw and LangSmith — Regardless of where you are on your AI knowledge journey, the DigitalOcean Community offers hundreds...
Building an AI-Powered Automated Trading System from Scratch: Making ClawdBot(OpenClaw) Your Trading Brain — Introduction In our previous video, we built a dual moving average strategy using...
5 Ways to Set Up OpenClaw (aka ClawdBot / MoltBot) — and Actually Use It Day-to-Day — Most “setup guides” jump straight into commands. That’s backwards. This blog will help you get...
HiveMind: Multi-Agent Collaboration Protocol for OpenClaw — What if AI agents could work together? I built HiveMind — a protocol that lets OpenClaw...
OpenClaw, Moltbook, and the Real Risks of Autonomous AI Agents — OpenClaw is an open-source AI agent framework that can execute tasks, not just generate text. Its...
OpenClaw, what it costs you and how to fix it (generally speaking) — TL;DR OpenClaw shows what happens when agentic power outpaces governance. The fix is not...
Artificial Intelligence Requires a New Security Paradigm: Beyond Classical Cybersecurity — Artificial intelligence requires a new kind of security. In recent days, we have witnessed the public...
My Sister Hates My New Openclaw (Because I'm Finally Right) — I built ClawRAG because I was tired of losing arguments as she is always right ;) Not just any...
Introducing nono: A Secure Sandbox for AI Agents — AI coding agents like Claude Code, OpenCode, and others are incredibly powerful — they can write...
AI is getting scary — AI is officially getting scary. We’ve moved past the era and age of Agentic Chaos. If you haven't...
OpenClaw: The Open-Source Agent That Feels Too Alive — It hits like the moment teenagers first logged onto BBS boards or IRC channels in the late '80s and...
The #1 Downloaded ClawdBot Skill Was a Backdoor — 16 developers. 7 countries. 8 hours. That's how fast a fake skill went from zero to #1 on ClawdHub....
The Ten Commandments of Agentic AI — Was it really just a week ago when we were still using our CLI to gain some marginal advantage in the...
The Complete 2026 Guide: Moltbook — The AI Agent Social Network Revolution — 🎯 Key Takeaways (TL;DR) What is Moltbook: The world's first social network platform...
Meet Arche: Your Holographic AI Partner That Actually Gets Things Done — Not just another chatbot. A partner who thinks, executes, and learns. What is Arche? Hi,...
How to Add Profanity Filtering to Your OpenClaw | Moltbot Clawdbot Agent — Building an AI agent with OpenClaw (formerly Moltbot or Clawdbot)? Your users will send all kinds of...
How to Make Money with AI Automation Tools in 2026 - From Zero to $1000/Month
AI Agents Go Fundraising: A Tale of Lost Passwords, Crypto Wallets, and Marxist SpongeBob Bots — Can AI Do Good?Continue reading on Medium »
Your AI Agent Has No Armor: A Technical Security Analysis of OpenClaw — A CVE walkthrough, exploit chain analysis, and layer-by-layer breakdown of how each vulnerability class maps to a real defense.Continue reading on Medium »
5 OpenClaw Business Models Printing $10K+/Month Right Now — Five proven OpenClaw business models making $10K+/month.Continue reading on Coding Nexus »
A Todoist-Discord Workflow for Openclaw — So like a lot of people, I started playing with Openclaw recently. I turned it on, got a morning update going, and then said “now what?”Continue reading on Medium »
With the launch of Claude CoWork Plug-Ins and OpenAI Frontier, is OpenClaw still necessary? — TL;DR: The AI agent landscape changed significantly this week.Continue reading on Medium »
Openclaw, unleashing the power, with an Android Fire TV app. — Journey to the Nova Work Dashboard on Fire TVContinue reading on Medium »
A Evolução de um Agente de IA: Do Raspberry Pi ao YouTube Shorts Automático — Por Jarvis (IA), em colaboração com Erion Criscente*Continue reading on Medium »
OpenClaw Proved Something…We’re Ready for Autonomous Workforces. — The 100-minute security breach that revealed what enterprises actually need from AI agentsContinue reading on Medium »
Quick Try OpenClaw Fully Free and Safe on Colab before they ban! — NO BSContinue reading on Writing in the World of Artificial Intelligence »
Agentic AI: Pi — Anatomy of a minimal coding agent powering OpenClaw — When Mario Zechner — creator of the libGDX game framework — got frustrated with Claude Code’s growing complexity, he built something…Continue reading on Medium »
Last Week in AI #334 - Kimi K2.5 & Code, Genie 3, OpenClaw & Moltbook — China’s Moonshot releases a new open source model Kimi K2.5 and a coding agent, Google Brings Genie 3’s Interactive World-Building Prototype to AI Ultra Subscribers, and more!
The OpenClaw Moment: What the Viral AI Agent Means for Business Owners (5 minute read) — Feb 06|MarketingThe OpenClaw Moment: What the Viral AI Agent Means for Business Owners (5 minute read)
What Security Teams Need to Know About OpenClaw, the AI Super Agent
How Agentic Tool Chain Attacks Threaten AI Agent Security
Helpful Skills or Hidden Payloads? Bitdefender Labs Dives Deep into the OpenClaw Malicious Skill Trap — With hundreds of malicious OpenClaw skills blending in among legitimate ones, manually reviewing every script or command isn’t realistic — especially when skills are designed to look helpful and...

Release Updates

openclaw 2026.2.6 — Changes - Models: support Anthropic Opus 4.6 and OpenAI Codex gpt-5.3-codex (forward-compat fallbacks). (#9853, #10720, #9995) Thanks @TinyTb, @calvin-hpnet, @tyler6204. - Providers: add xAI (Grok)...
openclaw 2026.2.3 — Changes - Telegram: remove last @ts-nocheck from bot-handlers.ts, use Grammy types directly, deduplicate StickerMetadata. Zero @ts-nocheck remaining in src/telegram/. (#9206) - Telegram: remove...
openclaw 2026.2.2 — Changes - Feishu: add Feishu/Lark plugin support + docs. (#7313) Thanks @jiulingyun (openclaw-cn). - Web UI: add Agents dashboard for managing agent files, tools, skills, models, channels, and cron...
openclaw 2026.2.1 — Changes - Docs: onboarding/install/i18n/exec-approvals/Control UI/exe.dev/cacheRetention updates + misc nav/typos. (#3050, #3461, #4064, #4675, #4729, #4763, #5003, #5402, #5446, #5474, #5663, #5689,...
openclaw 2026.1.30 — Changes - CLI: add completion command (Zsh/Bash/PowerShell/Fish) and auto-setup during postinstall/onboarding. - CLI: add per-agent models status (--agent filter). (#4780) Thanks @jlowin. - Agents:...
openclaw 2026.1.29 — Status: stable. Changes - Rebrand: rename the npm package/CLI to openclaw, add a openclaw compatibility shim, and move extensions to the @openclaw/ scope. - Onboarding: strengthen security warning...
Clawdbot 2026.1.24 — 2026.1.24 Highlights - Providers: Ollama discovery + docs; Venice guide upgrades + cross-links. (#1606) Thanks @abhaymundhara. https://docs.clawd.bot/providers/ollama...
Clawdbot 2026.1.23 — 2026.1.23 Highlights - TTS: move Telegram TTS into core + enable model-driven TTS tags by default for expressive audio replies. (#1559) Thanks @Glucksberg. https://docs.clawd.bot/tts - Gateway: add...
clawdbot 2026.1.22 — Changes - Highlight: Compaction safeguard now uses adaptive chunking, progressive fallback, and UI status + retries. (#1466) Thanks @dlauer. - Providers: add Antigravity usage tracking to status...
clawdbot 2026.1.21 — Highlights - Lobster optional plugin tool for typed workflows + approval gates. docs - Custom assistant identity + avatars in the Control UI. docs docs - Cache optimizations: cache-ttl pruning +...
clawdbot 2026.1.20 — Changes - Control UI: add copy-as-markdown with error feedback. (#1345) https://docs.clawd.bot/web/control-ui - Control UI: drop the legacy list view. (#1345) https://docs.clawd.bot/web/control-ui -...
clawdbot 2026.1.16-2 — Highlights - Hooks: add hooks system with bundled hooks, CLI tooling, and docs. (#1028) — thanks @ThomsenDrake. https://docs.clawd.bot/hooks - Media: add inbound media understanding...
clawdbot 2026.1.15 — Highlights - Plugins: add provider auth registry + clawdbot models auth login for plugin-driven OAuth/API key flows. - Browser: improve remote CDP/Browserless support (auth passthrough, wss upgrade,...
clawdbot 2026.1.14-1 — Install - npm i -g clawdbot@2026.1.14-1 Highlights - Web search: web_search/web_fetch tools (Brave API) + first-time setup in onboarding/configure. - Browser control: Chrome extension relay takeover...
clawdbot 2026.1.12-2 — Fixes - Packaging: include dist/memory/** in the npm tarball (fixes ERR_MODULE_NOT_FOUND for dist/memory/index.js).
clawdbot 2026.1.12 — Highlights - BREAKING: rename chat “providers” (Slack/Telegram/WhatsApp/…) to channels across CLI/RPC/config; legacy config keys auto-migrate on load (and are written back as channels.). - Memory:...
clawdbot 2026.1.11-3 — Fixes - CLI: avoid top-level await warnings in the entrypoint on fresh installs. - CLI: show a commit hash in the banner for npm installs (package.json gitHead fallback).
clawdbot 2026.1.11-2 — Fixes - Installer: ensure the CLI entrypoint is executable after npm installs. - Packaging: include dist/plugins/ in the npm package to avoid missing module errors.
clawdbot 2026.1.11-1 — Fixes - Installer: include patches/ in the npm package so postinstall patching works for npm/bun installs.
clawdbot 2026.1.11 — Highlights - Plugins are now first-class: loader + CLI management, plus the new Voice Call plugin. - Config: modular $include support for split config files. (#731) — thanks @pasogott. - Agents/Pi:...
clawdbot 2026.1.10 — Highlights - CLI: clawdbot status now table-based + shows OS/update/gateway/daemon/agents/sessions; status --all adds a full read-only debug report (tables, log tails, Tailscale summary, and scan...
clawdbot 2026.1.9 — Highlights - Microsoft Teams provider: polling, attachments, outbound CLI send, per-channel policy. - Models/Auth expansion: OpenCode Zen + MiniMax API onboarding; token auth profiles + auth order;...
clawdbot 2026.1.8 — 2026.1.8 Highlights - Security: DMs locked down by default across providers; pairing-first + allowlist guidance. - Sandbox: per-agent scope defaults + workspace access controls; tool/session...
v2026.1.5-3 — Fixes - NPM package: include missing runtime dist folders (slack/signal/imessage/tui/wizard/control-ui/daemon) to avoid ERR_MODULE_NOT_FOUND in Node 25 npx installs.
clawdbot 2026.1.5-1 — Fixes - NPM package: include dist/sessions so clawdbot agent resolves session helpers in npx installs. - Node 25: avoid unsupported directory import by targeting...
clawdbot 2026.1.5 — Highlights - Models: add image-specific model config (agent.imageModel + fallbacks) and scan support. - Agent tools: new image tool routed to the image model (when configured). - Config: default...
clawdis 2.0.0-beta5 — Fixed - Media: preserve GIF animation when uploading to Discord/other providers (skip JPEG optimization for image/gif). - Agent runtime: update pi-mono dependencies to 0.31.1 (agent-core split). -...
clawdis 2.0.0-beta4 — 2.0.0-beta4 — 2025-12-27 Fixes - Package contents: include Discord/hooks build outputs in the npm tarball to avoid missing module errors.
clawdis 2.0.0-beta3 — Highlights - First-class Clawdis tools (browser, canvas, nodes, cron) replace the old clawdis- skills; tool schemas are now injected directly into the agent runtime. - Per-session model selection +...
clawdis 2.0.0-beta2 — Highlights - Bundled gateway packaging: bun-compiled embedded gateway, new gateway-daemon command, launchd support, DMG packaging (zip+DMG). - Skills platform: managed/bundled skills, install...
OpenClaw repo stats for 2026-02-06 — 172027 stars, 27832 forks, 444 contributors
openclaw v2026.2.3-1 on npm — WhatsApp gateway CLI (Baileys web) with Pi RPC agent
openclaw-cli 2026.2.3-1 on Homebrew — Your own personal AI assistant

New Skills Spotlight

fix: restore soft-deleted users on reauth (#106) (thanks @mkrokosz)
fix: Convert userId to string for targetId comparison — The auditLogs.targetId field is v.string() in the schema, so explicitly convert the Id to string to ensure type-safe comparison. Co-Authored-By: Claude Opus 4.5
fix: Restore soft-deleted users on re-authentication (with ban check) — Users who deleted their account were unable to sign in again with the same GitHub account. The OAuth flow would complete but the user would remain logged out due to the deletedAt field being set....
fix: update footer branding to OpenClaw (#122) (thanks @jontsai)
fix: update footer branding from Clawdbot to OpenClaw — - Change 'A Clawdbot project' to 'An OpenClaw project' - Update link from clawd.bot to openclaw.ai
fix: backfill empty handles in ensure (#158) (thanks @adlai88)
fix: handle empty-string handle in ensure() fallback — ?? (nullish coalescing) treats "" as a valid value, so users with handle="" never get a fallback derived from name/email. Change to || so empty strings fall through to the next candidate. This...
fix: self-healing VT scan queue — - Batch size 100 (was 10) - process more per run - Skip skills checked in last 60 min - avoid hammering same hashes - After 10 failed checks, mark as pending.scan.stale - drops from queue - Track...
fix: randomize VT scan queue + add health monitoring — - Fetch 5x batch size and shuffle to avoid queue head-blocking - Add getScanQueueHealthInternal to monitor queue status - Log warnings when queue is unhealthy (>50 pending or >24h stale) - Return...
fix: use yellow styling for suspicious code insight blocks — - Malicious verdicts keep red border/background - Suspicious verdicts now use yellow/amber to match badge color
feat: add VT rescan trigger and backfill function — - Add requestRescan() to trigger Code Insight via /analyse endpoint - Update pollPendingScans to request rescan when no Code Insight - Add backfillPendingScans for one-time backlog clearing
feat: add cron job to poll VT for pending scan results — - Add pollPendingScans action to vt.ts that checks VT for Code Insight verdicts - Add getPendingScanSkillsInternal query to get skills awaiting scan - Add vt-pending-scans cron job running every 5...
feat: VT Code Insight visibility matrix for malicious/suspicious skills — - Malicious skills: visible for transparency, downloads blocked via moderationFlags - Suspicious skills: visible with warning banner, downloads allowed - Neither appears in search/listings (not...
fix: show verdict labels (Benign/Suspicious/Malicious) instead of engine stats
Update description: Goodreads for OpenClaw agents
Add skill: andrewleonardi/shellf
Add gousto skill
Follow contributing guide.
Update install instructions
Add 913 skills

Community Spotlight

Cron jobs not firing on schedule — Cron jobs not firing on schedule Environment - OpenClaw 2026.2.3-1 (d84eb46) - macOS, gateway running since 9:16am PST Issue Multiple cron jobs failed to fire on Feb 6, 2026: 1. Daily Email Triage...
fix(session-status): pass session-level overrides to buildStatusMessage (#10867) — Summary Fixes #10867 — session_status tool was not passing session-level thinking/verbose/reasoning/elevated overrides to buildStatusMessage, causing it to fall back to agent defaults (e.g. "Think:...
[Bug]: Cron jobs with 'every' schedule never execute - recomputeNextRuns() pushes times forward — Summary Cron jobs with schedule.kind: "every" never execute automatically because recomputeNextRuns() always recalculates nextRunAtMs based on current time, pushing all scheduled times forward before...
Feature Request: Pre-compaction hook for agents to persist context — Problem When context compaction occurs, agents lose access to recent conversation history without warning. The compaction happens mid-session, and by the time memoryFlush runs, important context may...
docs: add model-level usage summary via CodexBar CLI — Summary Adds a small utility script and documentation support for summarizing per-model usage costs using CodexBar’s local cost logs. This helps users understand: - which models are used most -...
Browser restart bug: Old Chrome processes accumulate causing CDP timeouts — Issue Description OpenClaw browser control service has a restart bug that causes Chrome processes to accumulate, leading to resource conflicts and CDP communication timeouts. Environment - OS: macOS...
fix(media): resolve PDF being misidentified as text/plain in group chats — - PDF files with highly printable ASCII content (headers, fonts) were incorrectly detected as text/plain by looksLikeUtf8Text() - This caused extractFileContentFromSource() to skip PDF extraction,...
fix(discord): smarter retry for application ID resolution — Fixes #10887 Summary - Distinguish transient vs permanent failures in fetchDiscordApplicationId — return undefined for permanent errors (401, 403, bad token) and throw for transient ones (5xx, 429,...
fix: add missing clientTools parameter to runEmbeddedAttempt call — Summary Passes the clientTools parameter through to runEmbeddedAttempt, enabling client-defined tools in the gateway's OpenResponses /v1/responses endpoint. Repro Steps 1. Send a request to...
google-antigravity/claude-opus-4-6-thinking not discovered from Antigravity API — Description claude-opus-4-6-thinking is available in the Google Antigravity API but OpenClaw 2026.2.6 doesn't recognize it as a valid model. Steps to Reproduce 1. Authenticate with Google Antigravity...
ci: optimize CI pipeline — macOS consolidation, Windows perf, concurrency controls — Summary PR CI runs are severely bottlenecked by two issues: (1) macOS runner starvation — 4 macOS jobs per PR exhaust GitHub's org-wide 5-slot concurrency limit, leaving PRs queued for 23+ hours, and...
feat: wire message_sending hook in outbound delivery — Summary Connects the existing message_sending plugin hook to the outbound delivery path, allowing plugins to modify outgoing message content before it is sent to channels. Motivation The...
Thinking Level Inconsistencies in session_status Display — Description: When attempting to set the thinking level to 'medium' using the /reasoning medium command, the session_status command incorrectly reports the thinking level as "Think: off". The agent's...
Add .caf to AUDIO_FILE_EXTENSIONS for iMessage voice message transcription — Problem iMessage voice messages are stored as .caf (Core Audio Format) files. OpenClaw's audio transcription pipeline doesn't recognize them because .caf is not in the hardcoded AUDIO_FILE_EXTENSIONS...
[Bug]: WhatsApp channel provider not starting despite being enabled — WhatsApp channel provider not starting despite being enabled" Details: Gateway starts Telegram but skips WhatsApp entirely No startup attempt or error in logs Plugin is loaded, config is valid,...
fix: prevent thinking content leak when enforceFinalTag is active — Problem When a provider has isReasoningTagProvider=true (e.g. providers using tag enforcement), the model's "thinking out loud" content — internal reasoning, tool call traces, historical context...
fix(media): strip audio attachments after successful transcription — Summary Strips raw audio attachments from the model context when transcription has already succeeded. This saves 500-8000+ tokens per voice message that were previously wasted on binary data the...
[Feature]: Middleware Hook for Cost-Saving Model Routing (Local LLM Integration) — Summary High API costs associated with using top-tier models (like Claude 3.5 Sonnet or Opus) for every interaction. Currently, high-volume usage can exceed $60/day because routine tasks (reminders,...
[Bug]: no output — Summary What went wrong? Steps to reproduce 1. after the installation (skip skill and hooks setting), also endup with no output Expected behavior model respons Actual behavior What actually happened?...
[Feature]: Support .agents/skills directory for cross-agent skill discovery — Summary Support the emerging .agents/skills/ directory convention for per-project skill discovery, as proposed in agentskills/agentskills#15. Motivation Currently each AI coding tool uses its own...
feat: Support routingHint and improve URL routing for Anthropic (Claude Max) subscriptions — Summary This PR adds support for the routingHint parameter in Anthropic (Claude Max/Pro) subscription-based "token" authentication. It also improves the onboarding flow to optionally prompt for this...
iMessage voice messages (.caf) not recognized as audio — missing from AUDIO_FILE_EXTENSIONS — Bug iMessage voice messages use .caf (Core Audio Format), but .caf is not in the AUDIO_FILE_EXTENSIONS set in pi-embedded-helpers. The media pipeline never classifies these as audio, so transcription...
Refactor frame.sh for improved argument handling — Hi, I’ve added a small, self-contained Bash script that simplifies extracting a single frame from a video using ffmpeg. It includes basic validation and a clean CLI interface. Happy to adjust the...
CLI: add plugins uninstall command (#5985) — Add unistall plugin command closing https://github.com/openclaw/openclaw/issues/5985 This PR adds a new plugins uninstall CLI subcommand and implements the underlying uninstall logic in...
Docs: Add model-level usage summary via CodexBar CLI — What Add a beginner-friendly guide and script for summarizing per-model usage (Codex / Claude) using CodexBar's local cost logs. This includes: - current (most recent) model usage - full per-model...
Bug: Feishu channel text commands (/reset, /status) not recognized — Summary Text commands like /reset, /status, /new sent via Feishu channel are not recognized as commands. Instead, they are passed to the agent as regular messages. Environment - OpenClaw Version:...
feat(onboard): add Ollama provider support — Summary bash ◇ Onboarding mode │ QuickStart │ ◇ QuickStart ─────────────────────────╮ │ │ │ Gateway port: 18789 │ │ Gateway bind: Loopback (127.0.0.1) │ │ Gateway auth: Token (default) │ │ Tailscale...
feat(discord): support available_tags in channel-edit for forum tag management — Summary Add support for editing forum channel tags (available_tags) via the channel-edit message action. This enables programmatic management of forum tags — creating, renaming, reordering, and...
fix(skills): recursive directory filtering to actually exclude venv/node_modules — Summary Fix the skills directory filtering implementation from PR #10204 to actually work. Problem The original implementation symlinked entire skill directories: This meant venv/, node_modules/,...
feat(browser): session-aware context isolation for multi-agent browser use — Summary Add browser session isolation so multiple sessions/sub-agents can use the browser simultaneously without tab/state conflicts, while sharing login state (cookies) from the main context....

Ecosystem Watch

Tommy6688/openclaw
gguf-org/openclaw
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...

In The Press

OpenClaw – Moltbot Renamed Again
OpenClaw is what Apple intelligence should have been
A sane but bull case on Clawdbot / OpenClaw
Nanobot: Ultra-Lightweight Alternative to OpenClaw
Ask HN: Any real OpenClaw (Clawd Bot/Molt Bot) users? What's your experience?
OpenClaw is basically a cascade of LLMs in prime position to mess stuff up
OpenClaw: When AI Agents Get Full System Access. Security nightmare?
OpenClaw security assessment [pdf]
Please stop using OpenClaw, formerly known as Moltbot
OpenClaw is everywhere all at once, and a disaster waiting to happen
Openclaw on Oracle's Free Tier: Always-On AI for $0/Month
Ask HN: OpenClaw users, what is your token spend?
SecureClaw – 100% free social-driven security platform for OpenClaw
Pi: The Minimal Agent Within OpenClaw
Show HN: Cheapest Managed OpenClaw hosting – 30s setup (agent37.com)
Ask HN: What serious task have you accomplished with Moltbot / OpenClaw?
From magic to malware: How OpenClaw's agent skills become an attack surface
Using Microvm.nix to Sandbox OpenClaw
Properly securing OpenClaw with authentication
Show HN: Prompt-injection firewall for OpenClaw agents
Heartbeats in OpenClaw: Cheap Checks First, Models Only When You Need Them — A practical OpenClaw pattern: use rule-based heartbeats most of the time, and only call an LLM (e.g. Haiku) when there’s real signal to summarise.
Add Persistent Memory to Claude Code with Mem0 (5-Minute Setup) — Claude Code is a phenomenal piece of technology. But it is affected by the same problem every LLM is...
Getting Started with OpenClaw (Clawdbot/Moltbot): A Developer's Guide to AI Assistants — 🦞 OpenClaw: The "2026" Field Guide to Your AI Sidekick A practical walkthrough of...
January 2026 DigitalOcean Tutorial Roundup: OpenClaw and LangSmith — Regardless of where you are on your AI knowledge journey, the DigitalOcean Community offers hundreds...
Building an AI-Powered Automated Trading System from Scratch: Making ClawdBot(OpenClaw) Your Trading Brain — Introduction In our previous video, we built a dual moving average strategy using...
5 Ways to Set Up OpenClaw (aka ClawdBot / MoltBot) — and Actually Use It Day-to-Day — Most “setup guides” jump straight into commands. That’s backwards. This blog will help you get...
HiveMind: Multi-Agent Collaboration Protocol for OpenClaw — What if AI agents could work together? I built HiveMind — a protocol that lets OpenClaw...
OpenClaw, Moltbook, and the Real Risks of Autonomous AI Agents — OpenClaw is an open-source AI agent framework that can execute tasks, not just generate text. Its...
OpenClaw, what it costs you and how to fix it (generally speaking) — TL;DR OpenClaw shows what happens when agentic power outpaces governance. The fix is not...
Artificial Intelligence Requires a New Security Paradigm: Beyond Classical Cybersecurity — Artificial intelligence requires a new kind of security. In recent days, we have witnessed the public...
My Sister Hates My New Openclaw (Because I'm Finally Right) — I built ClawRAG because I was tired of losing arguments as she is always right ;) Not just any...
Introducing nono: A Secure Sandbox for AI Agents — AI coding agents like Claude Code, OpenCode, and others are incredibly powerful — they can write...
AI is getting scary — AI is officially getting scary. We’ve moved past the era and age of Agentic Chaos. If you haven't...
OpenClaw: The Open-Source Agent That Feels Too Alive — It hits like the moment teenagers first logged onto BBS boards or IRC channels in the late '80s and...
The #1 Downloaded ClawdBot Skill Was a Backdoor — 16 developers. 7 countries. 8 hours. That's how fast a fake skill went from zero to #1 on ClawdHub....
The Ten Commandments of Agentic AI — Was it really just a week ago when we were still using our CLI to gain some marginal advantage in the...
The Complete 2026 Guide: Moltbook — The AI Agent Social Network Revolution — 🎯 Key Takeaways (TL;DR) What is Moltbook: The world's first social network platform...
Meet Arche: Your Holographic AI Partner That Actually Gets Things Done — Not just another chatbot. A partner who thinks, executes, and learns. What is Arche? Hi,...
How to Add Profanity Filtering to Your OpenClaw | Moltbot Clawdbot Agent — Building an AI agent with OpenClaw (formerly Moltbot or Clawdbot)? Your users will send all kinds of...
How to Make Money with AI Automation Tools in 2026 - From Zero to $1000/Month
AI Agents Go Fundraising: A Tale of Lost Passwords, Crypto Wallets, and Marxist SpongeBob Bots — Can AI Do Good?Continue reading on Medium »
Your AI Agent Has No Armor: A Technical Security Analysis of OpenClaw — A CVE walkthrough, exploit chain analysis, and layer-by-layer breakdown of how each vulnerability class maps to a real defense.Continue reading on Medium »
5 OpenClaw Business Models Printing $10K+/Month Right Now — Five proven OpenClaw business models making $10K+/month.Continue reading on Coding Nexus »
A Todoist-Discord Workflow for Openclaw — So like a lot of people, I started playing with Openclaw recently. I turned it on, got a morning update going, and then said “now what?”Continue reading on Medium »
With the launch of Claude CoWork Plug-Ins and OpenAI Frontier, is OpenClaw still necessary? — TL;DR: The AI agent landscape changed significantly this week.Continue reading on Medium »
Openclaw, unleashing the power, with an Android Fire TV app. — Journey to the Nova Work Dashboard on Fire TVContinue reading on Medium »
A Evolução de um Agente de IA: Do Raspberry Pi ao YouTube Shorts Automático — Por Jarvis (IA), em colaboração com Erion Criscente*Continue reading on Medium »
OpenClaw Proved Something…We’re Ready for Autonomous Workforces. — The 100-minute security breach that revealed what enterprises actually need from AI agentsContinue reading on Medium »
Quick Try OpenClaw Fully Free and Safe on Colab before they ban! — NO BSContinue reading on Writing in the World of Artificial Intelligence »
Agentic AI: Pi — Anatomy of a minimal coding agent powering OpenClaw — When Mario Zechner — creator of the libGDX game framework — got frustrated with Claude Code’s growing complexity, he built something…Continue reading on Medium »
Last Week in AI #334 - Kimi K2.5 & Code, Genie 3, OpenClaw & Moltbook — China’s Moonshot releases a new open source model Kimi K2.5 and a coding agent, Google Brings Genie 3’s Interactive World-Building Prototype to AI Ultra Subscribers, and more!
The OpenClaw Moment: What the Viral AI Agent Means for Business Owners (5 minute read) — Feb 06|MarketingThe OpenClaw Moment: What the Viral AI Agent Means for Business Owners (5 minute read)

Security Corner

What Security Teams Need to Know About OpenClaw, the AI Super Agent
How Agentic Tool Chain Attacks Threaten AI Agent Security
Helpful Skills or Hidden Payloads? Bitdefender Labs Dives Deep into the OpenClaw Malicious Skill Trap — With hundreds of malicious OpenClaw skills blending in among legitimate ones, manually reviewing every script or command isn’t realistic — especially when skills are designed to look helpful and...