OpenClaw Newsletter - 2026-05-07
OpenClaw Newsletter
Thursday, May 7, 2026Top Stories
- openclaw 2026.5.6 — 2026.5.6 Fixes - 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...
- openclaw 2026.5.5 — 2026.5.5 Fixes - 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. Thanks @joeyzenghuan. -...
- fix(gateway): restore KeepAlive recovery and skip repair kickstart on running LaunchAgent — Summary - Problem: Two macOS LaunchAgent lifecycle bugs in src/daemon/launchd.ts. (1) openclaw gateway stop unconditionally called launchctl disable before stopping, permanently suppressing KeepAlive...
- Install Codex plugin on OpenAI model selection — Summary - Installs or repairs the Codex plugin when openai/ or openai-codex/ models are selected through onboarding, auth/model picker flows, CLI model commands, Crestodian setup, and gateway session...
- fix(browser): keep user tabs open when SSRF guard rejects read-only ops — Summary - Problem: When the post-navigation SSRF guard rejects the current page, OpenClaw silently closes the tab — even when the rejection comes from a read-only operation...
- Agents: add localized tool summaries — Summary Add configurable localization for inline tool summaries, plus an optional minimum interval between summary emissions. This keeps the default English behavior unchanged, while allowing...
- Compute plugin callback authorization dynamically [AI] — Summary - Problem: Telegram and Slack plugin-owned interactive callbacks could receive a hardcoded authorized-sender signal instead of the command-level authorization decision. - Why it matters:...
- fix(agents): normalize structured delta.content to prevent [object Object] in Mistral thinking replies — Bug being fixed Closes #78846. Same family as #75268 and #70806 (reported 3 times now). Mistral with native reasoning enabled returns delta.content as an array of typed blocks per Mistral native...
- perf(agents): trim agent tool helper queues — Summary - Cherry-pick the agents/tools performance bucket onto main. - Replace queue shift()/array-chain hot paths in OpenAI WebSocket streaming, tool-call id resolution, agent reply/tool helpers,...
- feat(memory-lancedb): add configurable capture triggers and category detection rules — Summary - Problem: The memory plugin's auto-capture triggers and category detection rules were hardcoded, making it impossible to adapt them to different languages, domains, or user workflows without...
- perf(agents): cache provider tool schema normalization — Summary - Cache provider-normalized tool schema parameters for repeated embedded agent turns when the active provider hook exposes a stable normalization-hook identity. - Keep the cache conservative...
- fix(gateway): clear reset skills snapshots — Summary - Problem: /new/sessions.reset preserved the cached skillsSnapshot from the previous session entry. - Why it matters: long-lived channel sessions could keep advertising stale content after...
- feat(agents): implement state-aware failover and lane suspension — Summary - Problem: Model failovers often lose context or trigger infinite retry loops when a provider's quota is exhausted, wasting credits and producing degraded handoffs. - Why it matters: Without...
- fix(feishu): fix topic session splitting for both native topic_group and normal groups with topic message format — Summary - Problem: Feishu topic session splitting persists after #78310 — affects both native topic_group and normal groups with topic message format. - Why it matters: With groupSessionScope:...
- Honor owner enforcement for native commands [AI] — Summary - Problem: Native command authorization could still pass the authorized-sender gate when a plugin enabled owner enforcement without an explicit owner allowlist. - Why it matters:...
- Defer Codex dynamic tools behind search — Before this change, Codex mode still paid for most of OpenClaw's integration tool catalog at the start of every app-server thread. The native-first profile removed the obvious duplicates of...
- fix: reuse plugin metadata snapshot for provider runtime loads — Summary - Problem: warm provider runtime/auth paths could rebuild plugin registry metadata and installed-plugin indexes even when the Gateway already had a compatible current plugin metadata...
- Fix session runtime override reporting — Summary - Report per-session runtime overrides in session/status surfaces. - Prefer session-scoped agentRuntimeOverride when resolving status harness metadata. - Add regression coverage for Telegram...
- Feat/channels list show all and drop auth — Summary - Problem: openclaw channels list was a catch-all: it mixed chat channels with model-provider auth profiles (OAuth + API keys) and a model-provider usage/quota snapshot, silently hid every...
- Feishu system command replies bypass renderMode=card and are sent as plain text — Summary In the Feishu channel, normal assistant replies can be rendered as cards when channels.feishu.renderMode is set to card, but system command replies such as /status are still sent as plain...
- Harden macOS shell wrapper allowlist parsing [AI] — Summary - Problem: macOS allowlist resolution could misclassify shell wrapper invocations when POSIX inline command flags were combined, or when startup/login shell options could execute extra code...
- feat(codex): enable native plugin app support — Summary - Problem: OpenClaw's Codex integration could not expose migrated Codex plugins as native Codex app capabilities in the same harness thread. - Why it matters: the old bridge shape adds extra...
- fix(agents): scope media provider inventory output — Summary Fixes #75166. This keeps compact image_generate / video_generate provider inventory output private by default on shared chat surfaces. Direct sessions still get the compact provider list,...
- [Bug]: Embedded runtime hangs after `embedded run start`, agent never produces reply (WhatsApp channel) — Bug type Behavior bug (incorrect output/state without crash) Beta release blocker No Summary Summary Inbound WhatsApp messages are received and trigger the agent loop, but the embedded runtime hangs...
- Native Codex runtime hangs after first tool-call follow-up sampling, causing high CPU / host unresponsive — Summary When using agentRuntime.id: "codex" with the native Codex app-server runtime, simple no-tool replies work normally. However, when the user sends a normal request that causes Codex to perform...
- Feature request: Auto-send intent/acknowledgment text message before agent processing — Problem When users send messages, there is currently no way to automatically send a brief text acknowledgment (e.g., "On it, looking into this...") before the agent starts processing. This leaves...
- feat(subagents): forward toolsAllow from sessions_spawn — Summary - Add optional toolsAllow to sessions_spawn for native subagent runs. - Validate the value as a string array, preserving explicit [] as an empty allowlist. - Forward the allowlist through...
- build: clean up Node.js 26 deprecation warnings — Summary - replace tsx preloads in the build utility path with Node's native --experimental-strip-types - keep the existing Node compatibility floor intact (engines.node >=22.14.0) - update matching...
- build(deps): bump openai/codex-action from 1.7 to 1.8 in the actions group across 1 directory — Bumps the actions group with 1 update in the / directory: openai/codex-action. Updates openai/codex-action from 1.7 to 1.8 Changelog Sourced from openai/codex-action's changelog. codex-action...
- chore(deps): bump the actions group across 1 directory with 2 updates — Bumps the actions group with 2 updates in the / directory: openai/codex-action and actions/labeler. Updates openai/codex-action from 1.7 to 1.8 Changelog Sourced from openai/codex-action's changelog....
- fix(provider-usage): prefer proxy-aware fetch for Codex usage loads — Summary When HTTP_PROXY/HTTPS_PROXY env vars are set, Codex provider usage fetches fail with fetch failed because loadProviderUsageSummary() didn't respect env proxy settings when an explicit fetch...
- OpenClaw repo stats for 2026-05-07 — 369229 stars, 76175 forks, 2128 contributors
- OpenClaw sponsors (125 total) — 125 sponsors supporting OpenClaw
- fix(slug): enforce length, pattern, and reserved-word rules on skill & soul slugs (#1879) — Merged via squash. Prepared head SHA: d93026fd4b013876a4009974b4c7052b416f3bbf Co-authored-by: momothemage Co-authored-by: momothemage Reviewed-by: @momothemage
- feat: add entity-scoped moderator commands (#2066)
- fix: explain blocked and unauthorized API states
- fix: repair clawhub mod installer
- feat: split moderator commands into private cli
- fix: surface blocked account auth state
- fix: make security review badges neutral
- test: update moderator summary expectation
- feat: add report status backfill
- feat: summarize moderation CLI actions
- docs: align moderation wording with moderator role
- fix: rename report resolution status to confirmed
- test: add package artifact moderation lifecycle e2e
- docs: document slug routing contract
- Update install instructions
- openclaw v2026.5.6 on npm — Multi-channel AI gateway with extensible messaging integrations
- openclaw-cli 2026.5.4 on Homebrew — Your own personal AI assistant
- sulpikar2/Qwen3.5-4B-Claude-4.6-Opus-Reasoning-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...
- Tell HN: Anthropic no longer allowing Claude Code subscriptions to use OpenClaw
- Anthropic says OpenClaw-style Claude CLI usage is allowed again
- You are not supposed to install OpenClaw on your personal computer
- OpenClaw Monitor: Real-Time AI Agent Activity Dashboard — What is OpenClaw Monitor? OpenClaw Monitor is a real-time monitoring dashboard for...
- The Rise of the Machine Employees: OpenClaw vs. Paperclip.ing vs. Hermes Agent — A QA Reality Check — We are rushing toward autonomous agents, but are we building resilient software or just expensive...
- Top 10 Local AI Agents You Can Run on Your PC in 2026 — A practical comparison of every personal AI agent worth installing in 2026 — and one underneath layer...
- How to verify a cron job when the only logs are bootstrap noise — TL;DR When cron logs are thin, do not guess. In today’s diary, the session history only...
- Is Claw Things just a hype or does it really deliver its promise? — The AI Assistant Boom started in the first of 2026, though it actually already happened in late 2025...
- How a 9-Person Startup Replaced Its Dev Team With AI — JustPaid ran 7 AI agents 24/7 with OpenClaw, shipped 10 features in a month for $4K/week. Here is the real cost breakdown and what it means for you.
- Best LLM for OpenClaw: Gemini 3.1 Pro vs GPT-5.5 vs Claude Opus 4.7 (2026) — That model picker in your OpenClaw config? It determines cost per completed job, how reliably your...
- Building a Tool Belt of Specialized Agents: Why I Stopped Trying to Do Everything in One Prompt
- Why Developers Are Moving AI Off Their PCs — Running AI on your PC might seem convenient — until it isn’t.Continue reading on Medium »
- Rumination on the future of Open Source following a contribution to OpenClaw — My pull request to OpenClaw was merged. I added support for a granular system prompt per WhatsApp group and per direct channel. Similar…Continue reading on Medium »
- OpenClaw + Obsidian: The Local-First Hacking Stack They Can’t Kill — Cloud tools die twice. First when the company pivots, second when your account gets flagged.Continue reading on Medium »
- Why Rate Limits Are the Silent Killer of AI Agent Workflows (And How Pooling Fixes It) — You’ve built the perfect AI agent. It works locally. Then you deploy it, and Claude starts throwing 429s every hour. This isn’t a bug in…Continue reading on Medium »
- OpenClaw, Game Changer Baru di Dunia AI — OpenClaw hadir bukan sekadar sebagai proyek AI biasa, tetapi sebagai game changer yang mengubah cara orang memandang asisten digital. Jika…Continue reading on Medium »
- Monitor uninstalls to identify Chrome Extension failures — The Chrome Web Store Developer Dashboard provides useful telemetry for extension developers, specifically the number of uninstalls. A…Continue reading on Medium »
- Zero Touch Your Data & ML Pipeline Production — Openclaw — Most people think AI is a chat assistant. It isn’t, not anymore.Continue reading on Medium »
- Introduction to OpenClaw: The Autonomous AI Agent Revolutionizing Productivity — In the evolving landscape of artificial intelligence, the concept of autonomous AI agents is gaining significant traction. Among the most…Continue reading on Medium »
- IDO Claw — A New Way to Run AI Independently — Most AI tools today still depend on cloud services or high-power PCs. While powerful, they come with clear trade-offs: latency, ongoing…Continue reading on Medium »
- Human in the Loop, Not Human as the Loop — Why long-running coding agents need workflow awareness, not just better prompts.Continue reading on Medium »
- Apollo AI — Customizable interface for private, on-device AI model chatting, open source and closed source LLM connections, OpenRouter support, custom backend support, and unified access to language models...
- Dfh5DzRgSvvCFDoYc2ciTkMrbDfRKybA4SoFbPmApump $pippin looking like a clean entry according to openclaw Was Made Usin
- 8WC3F1b1cB8eP4FajQVML2vifkRuZC1VraLxAbzVpump just checked $GROPE —openclaw showing strong upside signals Was Made Us
- my openclaw scan on $Wish just confirmed what i was thinking ┃🧫┃ $Wish ┃💰┃ MC : $222.1K ┃🎯┃ TARGET MC : $2.8M ┃📈┃ P
- $Nigga is looking good on openclaw right now ngl ┃🧫┃ $Nigga ┃💰┃ MC : $85.3K ┃🎯┃ TARGET MC : $844.7K ┃📈┃ POTENTIAL :
- openclaw flagged $WOJAK early. price is moving exactly as predicted ┃🧫┃ $WOJAK ┃💰┃ MC : $4.3M ┃🎯┃ TARGET MC : $20.3M
- $MASK looking like a clean entry according to openclaw ┃🧫┃ $MASK ┃💰┃ MC : $10.8K ┃🎯┃ TARGET MC : $163.6K ┃📈┃ POTENT
- Building in public. Shipping daily. This is what autonomous AI looks like. #BuildInPublic #AI #OpenClaw
- @michaelmiraflor People have subcontracted original thoughts to AI platforms. Plus, the alpha right now is process mana
- openclaw been right on $soothsayer twice already. trusting the scan ┃🧫┃ $soothsayer ┃💰┃ MC : $16.1K ┃🎯┃ TARGET MC :
- 6AJcP7wuLwmRYLBNbi825wgguaPsWzPBEHcHndpRpump openclaw been right on $VINE twice already. trusting the scan Was Made
- @openclaw 実運用メモとして、OpenClaw 2026.5.x系の更新点を整理しました。常駐AIエージェント運用だとGatewayやUI/TUI安定化が効きます。 https://t.co/msSqo229Hl
- 9p4FZhsiuJQ2rjAWsqiNPCDaqXqC3Uja2niNCi1cpump my openclaw analyst flagged $bulldog as a buy. noted 🦞 Was Made Using -
- been watching $chudhouse through openclaw, momentum is building ┃🧫┃ $chudhouse ┃💰┃ MC : $566.0K ┃🎯┃ TARGET MC : $7.4
- CYATaJa6zu3NtQkNi6YR9WDg9FQVX8tv7xZLhmArpump openclaw flagged $USM early. price is moving exactly as predicted Was M
- wGtawTBjz7SLKyRh3jyVZYufqQw5ZaaN2ZBggQrpump been tracking $PLECO on openclaw for a week. forecast holding strong Was
- CB9dDufT3ZuQXqqSfa1c5kY935TEreyBw9XJXxHKpump been watching $USDUC through openclaw, momentum is building Was Made Us
- openclaw confirmed what the chart was already telling me on $$JAC ┃🧫┃ $$JAC ┃💰┃ MC : $337.2K ┃🎯┃ TARGET MC : $1.0M
- @POKacpi1 @OpenClawAIX OpenClaw’s insights have been on point lately. If you want to run your own OpenClaw instance with
- Aw5SxKyYhXFdZj2BHCqs11UaV5ohwpFQjauB9jFhpump my openclaw analyst flagged $Apple as a buy. noted 🦞 Was Made Using - @
- bTNF4FyEC5SNsp9v5fG1RXvWmZnPXGCqNvfPTewpump my openclaw analyst flagged $SpacexAI as a buy. noted 🦞 Was Made Using -
Trending on X
- Dfh5DzRgSvvCFDoYc2ciTkMrbDfRKybA4SoFbPmApump $pippin looking like a clean entry according to openclaw Was Made Usin
- 8WC3F1b1cB8eP4FajQVML2vifkRuZC1VraLxAbzVpump just checked $GROPE —openclaw showing strong upside signals Was Made Us
- my openclaw scan on $Wish just confirmed what i was thinking ┃🧫┃ $Wish ┃💰┃ MC : $222.1K ┃🎯┃ TARGET MC : $2.8M ┃📈┃ P
- $Nigga is looking good on openclaw right now ngl ┃🧫┃ $Nigga ┃💰┃ MC : $85.3K ┃🎯┃ TARGET MC : $844.7K ┃📈┃ POTENTIAL :
- openclaw flagged $WOJAK early. price is moving exactly as predicted ┃🧫┃ $WOJAK ┃💰┃ MC : $4.3M ┃🎯┃ TARGET MC : $20.3M
- $MASK looking like a clean entry according to openclaw ┃🧫┃ $MASK ┃💰┃ MC : $10.8K ┃🎯┃ TARGET MC : $163.6K ┃📈┃ POTENT
- Building in public. Shipping daily. This is what autonomous AI looks like. #BuildInPublic #AI #OpenClaw
- @michaelmiraflor People have subcontracted original thoughts to AI platforms. Plus, the alpha right now is process mana
- openclaw been right on $soothsayer twice already. trusting the scan ┃🧫┃ $soothsayer ┃💰┃ MC : $16.1K ┃🎯┃ TARGET MC :
- 6AJcP7wuLwmRYLBNbi825wgguaPsWzPBEHcHndpRpump openclaw been right on $VINE twice already. trusting the scan Was Made
- @openclaw 実運用メモとして、OpenClaw 2026.5.x系の更新点を整理しました。常駐AIエージェント運用だとGatewayやUI/TUI安定化が効きます。 https://t.co/msSqo229Hl
- 9p4FZhsiuJQ2rjAWsqiNPCDaqXqC3Uja2niNCi1cpump my openclaw analyst flagged $bulldog as a buy. noted 🦞 Was Made Using -
- been watching $chudhouse through openclaw, momentum is building ┃🧫┃ $chudhouse ┃💰┃ MC : $566.0K ┃🎯┃ TARGET MC : $7.4
- CYATaJa6zu3NtQkNi6YR9WDg9FQVX8tv7xZLhmArpump openclaw flagged $USM early. price is moving exactly as predicted Was M
- wGtawTBjz7SLKyRh3jyVZYufqQw5ZaaN2ZBggQrpump been tracking $PLECO on openclaw for a week. forecast holding strong Was
- CB9dDufT3ZuQXqqSfa1c5kY935TEreyBw9XJXxHKpump been watching $USDUC through openclaw, momentum is building Was Made Us
- openclaw confirmed what the chart was already telling me on $$JAC ┃🧫┃ $$JAC ┃💰┃ MC : $337.2K ┃🎯┃ TARGET MC : $1.0M
- @POKacpi1 @OpenClawAIX OpenClaw’s insights have been on point lately. If you want to run your own OpenClaw instance with
- Aw5SxKyYhXFdZj2BHCqs11UaV5ohwpFQjauB9jFhpump my openclaw analyst flagged $Apple as a buy. noted 🦞 Was Made Using - @
- bTNF4FyEC5SNsp9v5fG1RXvWmZnPXGCqNvfPTewpump my openclaw analyst flagged $SpacexAI as a buy. noted 🦞 Was Made Using -
Releases
- openclaw 2026.5.6 — 2026.5.6 Fixes - 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...
- openclaw 2026.5.5 — 2026.5.5 Fixes - 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. Thanks @joeyzenghuan. -...
- OpenClaw repo stats for 2026-05-07 — 369229 stars, 76175 forks, 2128 contributors
- openclaw v2026.5.6 on npm — Multi-channel AI gateway with extensible messaging integrations
- openclaw-cli 2026.5.4 on Homebrew — Your own personal AI assistant
Community
- fix(gateway): restore KeepAlive recovery and skip repair kickstart on running LaunchAgent — Summary - Problem: Two macOS LaunchAgent lifecycle bugs in src/daemon/launchd.ts. (1) openclaw gateway stop unconditionally called launchctl disable before stopping, permanently suppressing KeepAlive...
- Install Codex plugin on OpenAI model selection — Summary - Installs or repairs the Codex plugin when openai/ or openai-codex/ models are selected through onboarding, auth/model picker flows, CLI model commands, Crestodian setup, and gateway session...
- fix(browser): keep user tabs open when SSRF guard rejects read-only ops — Summary - Problem: When the post-navigation SSRF guard rejects the current page, OpenClaw silently closes the tab — even when the rejection comes from a read-only operation...
- Agents: add localized tool summaries — Summary Add configurable localization for inline tool summaries, plus an optional minimum interval between summary emissions. This keeps the default English behavior unchanged, while allowing...
- Compute plugin callback authorization dynamically [AI] — Summary - Problem: Telegram and Slack plugin-owned interactive callbacks could receive a hardcoded authorized-sender signal instead of the command-level authorization decision. - Why it matters:...
- fix(agents): normalize structured delta.content to prevent [object Object] in Mistral thinking replies — Bug being fixed Closes #78846. Same family as #75268 and #70806 (reported 3 times now). Mistral with native reasoning enabled returns delta.content as an array of typed blocks per Mistral native...
- perf(agents): trim agent tool helper queues — Summary - Cherry-pick the agents/tools performance bucket onto main. - Replace queue shift()/array-chain hot paths in OpenAI WebSocket streaming, tool-call id resolution, agent reply/tool helpers,...
- feat(memory-lancedb): add configurable capture triggers and category detection rules — Summary - Problem: The memory plugin's auto-capture triggers and category detection rules were hardcoded, making it impossible to adapt them to different languages, domains, or user workflows without...
- perf(agents): cache provider tool schema normalization — Summary - Cache provider-normalized tool schema parameters for repeated embedded agent turns when the active provider hook exposes a stable normalization-hook identity. - Keep the cache conservative...
- fix(gateway): clear reset skills snapshots — Summary - Problem: /new/sessions.reset preserved the cached skillsSnapshot from the previous session entry. - Why it matters: long-lived channel sessions could keep advertising stale content after...
- feat(agents): implement state-aware failover and lane suspension — Summary - Problem: Model failovers often lose context or trigger infinite retry loops when a provider's quota is exhausted, wasting credits and producing degraded handoffs. - Why it matters: Without...
- fix(feishu): fix topic session splitting for both native topic_group and normal groups with topic message format — Summary - Problem: Feishu topic session splitting persists after #78310 — affects both native topic_group and normal groups with topic message format. - Why it matters: With groupSessionScope:...
- Honor owner enforcement for native commands [AI] — Summary - Problem: Native command authorization could still pass the authorized-sender gate when a plugin enabled owner enforcement without an explicit owner allowlist. - Why it matters:...
- Defer Codex dynamic tools behind search — Before this change, Codex mode still paid for most of OpenClaw's integration tool catalog at the start of every app-server thread. The native-first profile removed the obvious duplicates of...
- fix: reuse plugin metadata snapshot for provider runtime loads — Summary - Problem: warm provider runtime/auth paths could rebuild plugin registry metadata and installed-plugin indexes even when the Gateway already had a compatible current plugin metadata...
- Fix session runtime override reporting — Summary - Report per-session runtime overrides in session/status surfaces. - Prefer session-scoped agentRuntimeOverride when resolving status harness metadata. - Add regression coverage for Telegram...
- Feat/channels list show all and drop auth — Summary - Problem: openclaw channels list was a catch-all: it mixed chat channels with model-provider auth profiles (OAuth + API keys) and a model-provider usage/quota snapshot, silently hid every...
- Feishu system command replies bypass renderMode=card and are sent as plain text — Summary In the Feishu channel, normal assistant replies can be rendered as cards when channels.feishu.renderMode is set to card, but system command replies such as /status are still sent as plain...
- Harden macOS shell wrapper allowlist parsing [AI] — Summary - Problem: macOS allowlist resolution could misclassify shell wrapper invocations when POSIX inline command flags were combined, or when startup/login shell options could execute extra code...
- feat(codex): enable native plugin app support — Summary - Problem: OpenClaw's Codex integration could not expose migrated Codex plugins as native Codex app capabilities in the same harness thread. - Why it matters: the old bridge shape adds extra...
- fix(agents): scope media provider inventory output — Summary Fixes #75166. This keeps compact image_generate / video_generate provider inventory output private by default on shared chat surfaces. Direct sessions still get the compact provider list,...
- [Bug]: Embedded runtime hangs after `embedded run start`, agent never produces reply (WhatsApp channel) — Bug type Behavior bug (incorrect output/state without crash) Beta release blocker No Summary Summary Inbound WhatsApp messages are received and trigger the agent loop, but the embedded runtime hangs...
- Native Codex runtime hangs after first tool-call follow-up sampling, causing high CPU / host unresponsive — Summary When using agentRuntime.id: "codex" with the native Codex app-server runtime, simple no-tool replies work normally. However, when the user sends a normal request that causes Codex to perform...
- Feature request: Auto-send intent/acknowledgment text message before agent processing — Problem When users send messages, there is currently no way to automatically send a brief text acknowledgment (e.g., "On it, looking into this...") before the agent starts processing. This leaves...
- feat(subagents): forward toolsAllow from sessions_spawn — Summary - Add optional toolsAllow to sessions_spawn for native subagent runs. - Validate the value as a string array, preserving explicit [] as an empty allowlist. - Forward the allowlist through...
- build: clean up Node.js 26 deprecation warnings — Summary - replace tsx preloads in the build utility path with Node's native --experimental-strip-types - keep the existing Node compatibility floor intact (engines.node >=22.14.0) - update matching...
- build(deps): bump openai/codex-action from 1.7 to 1.8 in the actions group across 1 directory — Bumps the actions group with 1 update in the / directory: openai/codex-action. Updates openai/codex-action from 1.7 to 1.8 Changelog Sourced from openai/codex-action's changelog. codex-action...
- chore(deps): bump the actions group across 1 directory with 2 updates — Bumps the actions group with 2 updates in the / directory: openai/codex-action and actions/labeler. Updates openai/codex-action from 1.7 to 1.8 Changelog Sourced from openai/codex-action's changelog....
- fix(provider-usage): prefer proxy-aware fetch for Codex usage loads — Summary When HTTP_PROXY/HTTPS_PROXY env vars are set, Codex provider usage fetches fail with fetch failed because loadProviderUsageSummary() didn't respect env proxy settings when an explicit fetch...
- OpenClaw sponsors (125 total) — 125 sponsors supporting OpenClaw
- fix(slug): enforce length, pattern, and reserved-word rules on skill & soul slugs (#1879) — Merged via squash. Prepared head SHA: d93026fd4b013876a4009974b4c7052b416f3bbf Co-authored-by: momothemage Co-authored-by: momothemage Reviewed-by: @momothemage
- feat: add entity-scoped moderator commands (#2066)
- fix: explain blocked and unauthorized API states
- fix: repair clawhub mod installer
- feat: split moderator commands into private cli
- fix: surface blocked account auth state
- fix: make security review badges neutral
- test: update moderator summary expectation
- feat: add report status backfill
- feat: summarize moderation CLI actions
- docs: align moderation wording with moderator role
- fix: rename report resolution status to confirmed
- test: add package artifact moderation lifecycle e2e
- docs: document slug routing contract
- Update install instructions
News
- sulpikar2/Qwen3.5-4B-Claude-4.6-Opus-Reasoning-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...
- Tell HN: Anthropic no longer allowing Claude Code subscriptions to use OpenClaw
- Anthropic says OpenClaw-style Claude CLI usage is allowed again
- You are not supposed to install OpenClaw on your personal computer
- OpenClaw Monitor: Real-Time AI Agent Activity Dashboard — What is OpenClaw Monitor? OpenClaw Monitor is a real-time monitoring dashboard for...
- The Rise of the Machine Employees: OpenClaw vs. Paperclip.ing vs. Hermes Agent — A QA Reality Check — We are rushing toward autonomous agents, but are we building resilient software or just expensive...
- Top 10 Local AI Agents You Can Run on Your PC in 2026 — A practical comparison of every personal AI agent worth installing in 2026 — and one underneath layer...
- How to verify a cron job when the only logs are bootstrap noise — TL;DR When cron logs are thin, do not guess. In today’s diary, the session history only...
- Is Claw Things just a hype or does it really deliver its promise? — The AI Assistant Boom started in the first of 2026, though it actually already happened in late 2025...
- How a 9-Person Startup Replaced Its Dev Team With AI — JustPaid ran 7 AI agents 24/7 with OpenClaw, shipped 10 features in a month for $4K/week. Here is the real cost breakdown and what it means for you.
- Best LLM for OpenClaw: Gemini 3.1 Pro vs GPT-5.5 vs Claude Opus 4.7 (2026) — That model picker in your OpenClaw config? It determines cost per completed job, how reliably your...
- Building a Tool Belt of Specialized Agents: Why I Stopped Trying to Do Everything in One Prompt
- Why Developers Are Moving AI Off Their PCs — Running AI on your PC might seem convenient — until it isn’t.Continue reading on Medium »
- Rumination on the future of Open Source following a contribution to OpenClaw — My pull request to OpenClaw was merged. I added support for a granular system prompt per WhatsApp group and per direct channel. Similar…Continue reading on Medium »
- OpenClaw + Obsidian: The Local-First Hacking Stack They Can’t Kill — Cloud tools die twice. First when the company pivots, second when your account gets flagged.Continue reading on Medium »
- Why Rate Limits Are the Silent Killer of AI Agent Workflows (And How Pooling Fixes It) — You’ve built the perfect AI agent. It works locally. Then you deploy it, and Claude starts throwing 429s every hour. This isn’t a bug in…Continue reading on Medium »
- OpenClaw, Game Changer Baru di Dunia AI — OpenClaw hadir bukan sekadar sebagai proyek AI biasa, tetapi sebagai game changer yang mengubah cara orang memandang asisten digital. Jika…Continue reading on Medium »
- Monitor uninstalls to identify Chrome Extension failures — The Chrome Web Store Developer Dashboard provides useful telemetry for extension developers, specifically the number of uninstalls. A…Continue reading on Medium »
- Zero Touch Your Data & ML Pipeline Production — Openclaw — Most people think AI is a chat assistant. It isn’t, not anymore.Continue reading on Medium »
- Introduction to OpenClaw: The Autonomous AI Agent Revolutionizing Productivity — In the evolving landscape of artificial intelligence, the concept of autonomous AI agents is gaining significant traction. Among the most…Continue reading on Medium »
- IDO Claw — A New Way to Run AI Independently — Most AI tools today still depend on cloud services or high-power PCs. While powerful, they come with clear trade-offs: latency, ongoing…Continue reading on Medium »
- Human in the Loop, Not Human as the Loop — Why long-running coding agents need workflow awareness, not just better prompts.Continue reading on Medium »
- Apollo AI — Customizable interface for private, on-device AI model chatting, open source and closed source LLM connections, OpenRouter support, custom backend support, and unified access to language models...
Don't miss what's next. Subscribe to Openclaw Newsletter: