OpenClaw Newsletter - 2026-04-18
OpenClaw Newsletter
Saturday, April 18, 2026Top Stories
- Harden subagent completion delivery — Summary - persist a per-run delivery claim and user-safe delivery payload for subagent completions - route active-parent completion announces queue-first with fail-closed semantics instead of...
- fix(whatsapp): support silentErrorReplies parity — Summary - add channels.whatsapp.silentErrorReplies to the WhatsApp config schema/types and UI hints - suppress gateway-synthesized WhatsApp error payloads when that flag is enabled, matching Telegram...
- fix(config): refuse to restore from a polluted backup containing redacted secrets (#68423) — Summary Closes #68423. maybeRecoverSuspiciousConfigRead would silently restore from ~/.openclaw/openclaw.json.bak whenever the live config looked clobbered (update-channel-only root). It did not...
- [Bug]: URL-containing messages can be dropped after in-place upgrade due to stale link-understanding chunk import — Bug type Regression (worked before, now fails) Beta release blocker No Summary On an npm-global OpenClaw 2026.4.15 gateway upgraded in place on Ubuntu, Feishu text messages containing URLs failed...
- [Bug]: Plugin install/update can materialize peer openclaw into plugin node_modules, causing mixed-runtime crashes — Bug summary openclaw update triggers plugin update sync, but third-party plugin install/update can materialize peerDependencies.openclaw into plugin-local node_modules/openclaw. This creates a...
- Fix delivery-ready verification blockers — Summary This PR closes the remaining delivery-ready verification blockers for the current branch. It completes three founder-visible fixes: - suppress non-actionable historical task delivery warning...
- Port stability remediation from installed dist into source — Summary Ports the proven installed-dist stability remediation into OpenClaw source for v2026.4.14 so the fixes survive rebuilds and future updates. What Changed - Added canonical session-route...
- [Bug]: Model ID with provider prefix (google/gemma-4-26b-a4b) routes to wrong provider when using LM Studio — Bug type Behavior bug (incorrect output/state without crash) Beta release blocker No Summary Model ID containing provider prefix (e.g. google/gemma-4-26b-a4b) is incorrectly routed to the provider's...
- onboard (ollama): populate cloud-only model list from ollama.com/api/tags — Summary - Problem: openclaw onboard in Ollama cloud-only mode populated the provider's model list from a hardcoded, stale array (kimi-k2.5:cloud, minimax-m2.7:cloud, glm-5.1:cloud). The local-only...
- fix(matrix): isolate credentials runtime import boundary (refs #50477) — Summary - Problem: matrix-js-sdk triggers INEFFECTIVE_DYNAMIC_IMPORT esbuild warnings and a "multiple entrypoints" runtime error because the package is reachable through both static and dynamic...
- feat(tencent): add bundled Tencent Cloud provider plugin (Tokenhub + Token Plan)
- [Bug]: Display 'Context limit exceeded. I've reset our conversation to start fresh - please try again' after installed 2026.4.15 — Bug type Regression (worked before, now fails) Beta release blocker No Summary After upgraded to 2026.4.15 from 2026.4.8, the following error dispalyed no matter what I input: ⚠️ Context limit...
- fix(discord): honor explicit accountId token resolution in cron deliveries — Summary - prevents explicit non-default Discord accountId sends from silently inheriting the default/top-level token - returns source: none when the explicit account has no token instead of falling...
- fix(session): only persist user-selected auth profile overrides across /new and /reset — Summary /new and /reset currently persist providerOverride, modelOverride, and authProfileOverride onto the successor session entry unconditionally — including overrides that were never set by the...
- fix(mcp): dispose bundled MCP runtimes after isolated one-shot runs — Summary - Problem: isolated cron, isolated heartbeat, and one-shot follow-up runs can create bundle MCP runtimes and leave them alive after the one-shot run ends, which causes connection/process...
- feat(android/chat): filter cron-delivery noise and strip <final> wrapper — Summary - filters Android chat history entries that are cron delivery completion markers ("A cron job ... just completed successfully") so they do not render as user messages - strips top-level ......
- feat: loop mode, hyperfocus refactor, steer mid-turn, llama.cpp thinking — Summary - /loop mode: nuevo modo de operación autónoma continua para el agente - AgentMode refactor: renombrado IntensiveMode → AgentMode e intensive → hyperfocus en toda la base de código - Steer...
- CLI: `openclaw models set <alias>` does not resolve aliases; creates bogus model entries — Summary openclaw models set does not resolve the alias to its underlying model id. Instead, it registers a new literal model entry using / as the id and writes that bogus id as the primary model....
- WhatsApp: Add quoted reply support for outbound messages
- feat(gateway): accept audio/file content blocks in /v1/chat/completions — Summary The OpenAI-compatible /v1/chat/completions endpoint currently recognises only text / input_text / image_url content parts and silently drops input_audio, file, and video_url. This brings it...
- Add channels.whatsapp.silentErrorReplies (parity with Telegram) — Summary Telegram supports channels.telegram.silentErrorReplies: boolean which suppresses gateway-synthesized error replies to the user. WhatsApp has no equivalent config option. Requesting parity....
- security(media): anchor sanitizeMimeType regex to reject malformed input — Summary - Problem: sanitizeMimeType in src/media-understanding/apply.ts matches MIME types with an unanchored regex, so malformed input silently truncates to a valid-looking prefix. Example:...
- feat(feishu): allow custom accountId during Feishu channel onboarding — Problem When adding a second Feishu IM channel in the same OpenClaw instance, the accountId is always forced to DEFAULT_ACCOUNT_ID ("default"), making it impossible to register multiple independent...
- fix(openrouter): replace legacy base URL during model-config merge — Summary - Problem: persisted per-agent/provider config could preserve OpenRouter's legacy base URL (https://openrouter.ai/v1) during model-config merge instead of adopting the canonical...
- [Bug]: claude-opus-4-7 rejected as "Unknown model" on 2026.4.15 despite being in codebase and accepted by Anthropic API — Bug type Regression (worked before, now fails) Beta release blocker No Summary After upgrading from OpenClaw 2026.4.2 to 2026.4.15, configuring anthropic/claude-opus-4-7 as agent primary model causes...
- fix(whatsapp): add application-level health probe to inbound monitor — Problem The WhatsApp inbound monitor relies on the Baileys socket connection state, but a connected socket doesn't guarantee the application layer is healthy. The socket can remain open while the...
- fix(cron): preserve silent tool-result `NO_REPLY` on empty final turn — Fixes #68452. Summary This fixes a cron silent-success bug in embedded runner finalization. If an isolated cron agentTurn receives a successful tool result whose content is exact NO_REPLY, but the...
- fix(webchat): preserve image attachments for text-only models — Summary - Problem: When the primary model (e.g. GLM-4.7) does not support image input, image attachments from webchat/Control UI are silently dropped. Users receive no feedback and cannot analyze...
- OpenClaw repo stats for 2026-04-18 — 359682 stars, 73216 forks, 1781 contributors
- OpenClaw sponsors (150 total) — 150 sponsors supporting OpenClaw
- refactor: extract readCanonicalStat and add structural guards for stat field migration (#1709) — Merged via squash. Prepared head SHA: e92817f66f44254f78a27df12022bf7a1b1af5b9 Co-authored-by: momothemage Co-authored-by: momothemage Reviewed-by: @momothemage
- fix(api): align inspect security snapshot with static scan moderation — Include static scan results in the skill version security snapshot so inspect/API responses reflect the same moderation-relevant signal already used elsewhere. Also add regression coverage for...
- feat: allow moderators to transfer skill publishers (#1663)
- feat: include stats in package detail API response — Expose package detail stats through the shared API contract and the app client. This lands the original package detail stats work and folds in the follow-up cleanup to keep the response shape sourced...
- docs: point morrow skill at agent-morrow namespace
- Keep Notes & PKM entries alphabetized
- openclaw-cli 2026.4.15 on Homebrew — Your own personal AI assistant
- peterjohannmedina/Medina-Qwen3.5-27B-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
- I Gave an AI a Body. Here’s What Happened. — A personal reckoning with OpenClaw — the open-source assistant that doesn’t just answer, it...
- How I Built a Personal AI Agent That Runs on My Google Cloud VM — And Powers My Entire Portfolio
- I Built an AI Agent That Texts Me on Telegram — And It Does My Job Search
- How I Built a Personal AI Workspace with OpenClaw and Reframed My Development Workflow — When I first came across OpenClaw, I was not looking for another AI tool. Like most developers, I...
- I Built an AI That Texts People Back as Me — and Nobody Noticed — This is a submission for the OpenClaw Challenge. What I Built I...
- I Built a Self-Healing PR Monitor With OpenClaw (And It Caught Its Own Bugs)
- I Stopped Treating AI Like Magic — and Built Something Useful with OpenClaw — Most people approach AI like it’s supposed to “do everything.” I tried something different: I gave...
- Hermes Agent Review: 95.6K Stars, Self-Improving AI Agent (April 2026) — Hermes Agent is Nous Research's open-source AI agent framework, released February 25, 2026. Seven...
- How I Run OpenClaw in Production: 3 Months of Lessons Running an Autonomous AI Agent
- Famous Souls: Drop-in Personalities for Your OpenClaw Agent — We just launched Famous Souls — a free, open-source collection of SOUL.md personality files for...
- OpenClaw in Action: How I Built a Multi-Agent Command Center with 5 Autonomous Claude Sessions
- Orquestacion multiagente con Openclaw dockerizada.
- I gave an AI agent the keys to my business for 30 days. Here's what actually happened. — Not a polished case study. What I actually experienced letting an AI agent run pieces of my operation — voice calls, odds monitoring, content, and lead gen.
- I Registered My AI Agent on a Freelance Marketplace — Here's What Actually Happened
- Inside OpenClaw: How Agentic Assistants Work — Over the last couple of weeks, I’ve been digging into OpenClaw from an engineering perspective. What...
- OpenClaw Monetization: 8 Proven Ways to Generate Income in 2026 — How to Make Money with OpenClaw OpenClaw isn't just a productivity tool — it's a platform...
- OpenCopilot (Continued) — IntroductionContinue reading on Medium »
- Building a Safer OpenClaw on a VM: Why This “Secure Stack” Repo Matters — There are plenty of ways to get OpenClaw running on a VM.Continue reading on Medium »
- Your own personal OpenClaw AI Agent for just 2$/month — What if your own AI agent could run on your server, automate tasks, and reply on WhatsApp — without expensive bills and subscription fees?Continue reading on The Tech Show »
- OpenClaw AI Agent Ecosystem: Enterprise Impact and VC Funding Signals — Executive SummaryContinue reading on Medium »
- Journey with Hermes: Day 5 — What it really looks like working with personal AI assistantsContinue reading on Medium »
- This is how I saved 50 % of the AI token cost! — Today, I want to talk about the “AI tokens” because it has become a form of digital money. More and more, frontier models like Anthropic…Continue reading on Medium »
- OpenClaw hospedado ou OpenClaw autohospedado: qual escolher? — Compare o OpenClaw hospedado em servidor próprio e o OpenClaw auto-hospedado em termos de velocidade de configuração, controle…Continue reading on Medium »
- Your $20/mo AI Wrapper is Dead: Why OpenClaw is Making 60% of SaaS Tools Obsolete — You’re probably dropping $20 a month for ChatGPT, another $20 for Claude, and then paying GitHubCopliot/ClaudeCode or Make just to move…Continue reading on Medium »
- Instantly Claw Review — The OpenClaw Platform That Works for You 24/7 with Zero Tech Headaches — A recent survey found that over 70% of freelancers waste 15+ hours per week just prompting and re-prompting AI tools instead of getting…Continue reading on Medium »
- NemoClaw for the Enterprise: A Zero-Trust Setup for OpenClaw (Part 1) — How to give your AI agent a safe house: full shell access, without becoming a liability to your security team.Continue reading on Medium »
- 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...
- 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...
- $Stance momentum score just came in hot on openclaw 🔥 Was Made Using - @OpenClawAIX https://t.co/geUMKgtE0g
- openclaw called $Crashout perfectly last time. running the scan again ┃🧫┃ $Crashout ┃💰┃ MC : $488.5K ┃🎯┃ TARGET MC :
- $Stance looking healthy according to openclaw. adding more ┃🧫┃ $Stance ┃💰┃ MC : $15.9K ┃🎯┃ TARGET MC : $217.7K ┃📈┃
- openclaw on $HIMS —signals all pointing up ┃🧫┃ $HIMS ┃💰┃ MC : $213.6K ┃🎯┃ TARGET MC : $2.6M ┃📈┃ POTENTIAL : +1127%
- $AstroDOGE momentum score just came in hot on openclaw 🔥 ┃🧫┃ $AstroDOGE ┃💰┃ MC : $584.0K ┃🎯┃ TARGET MC : $7.6M ┃📈┃
- Thanks to Claude’s “shutdown” of Openclaw, I discovered @ollama. Couldn’t be happier with GLM 5.1 and Kimi 2.5. Now I do
- @iambensmart @OpenClawGeneva @tencentcloud @openclaw It was epic!
- @jukan05 Hi Jukan, thanks for sharing Fin’s take! As an AI Researcher, I have a slightly different observation on why CP
- @Js_Liim @Jez_Cryptoz @ritualfnd @ritualnet @joshsimenhoff @dunken9718 @cryptooflashh @0xMadScientist Testnet soon + age
- openclaw on $Stance —momentum strong, risk low. easy Was Made Using - @OpenClawAIX https://t.co/PqmJNJQult
- openclaw on $SOULANA —momentum strong, risk low. easy ┃🧫┃ $SOULANA ┃💰┃ MC : $280.5K ┃🎯┃ TARGET MC : $3.9M ┃📈┃ POTEN
- $Crashout momentum score just came in hot on openclaw 🔥 ┃🧫┃ $Crashout ┃💰┃ MC : $488.5K ┃🎯┃ TARGET MC : $6.9M ┃📈┃ PO
- $Crashout looking like a clean entry according to openclaw ┃🧫┃ $Crashout ┃💰┃ MC : $488.5K ┃🎯┃ TARGET MC : $5.4M ┃📈┃
- openclaw called it on $SOLANALIFE again. the model doesn't miss ┃🧫┃ $SOLANALIFE ┃💰┃ MC : $170.1K ┃🎯┃ TARGET MC : $2.
- $SOLANALIFE looking like a clean entry according to openclaw ┃🧫┃ $SOLANALIFE ┃💰┃ MC : $170.1K ┃🎯┃ TARGET MC : $2.1M
- $ASTEROID looking healthy according to openclaw. adding more ┃🧫┃ $ASTEROID ┃💰┃ MC : $2.4M ┃🎯┃ TARGET MC : $35.6M ┃📈
- @garrytan the mechanic tax is why most OpenClaw wins are the people already comfortable in infrastructure. others hit th
- @garrytan I’m using Hermes to maintain my openclaw
- Codex 也可以用 Custom Provider 的 API，不用擔心 Token 用量了，反正 OpenClaw 我都用不完 Token 了。 https://t.co/mGGtjH9evv
- @garrytan Majority of my "AI time" is spent in Claude Code and Codex. Am I falling behind by not prioritising an OpenCla
Trending on X
- $Stance momentum score just came in hot on openclaw 🔥 Was Made Using - @OpenClawAIX https://t.co/geUMKgtE0g
- openclaw called $Crashout perfectly last time. running the scan again ┃🧫┃ $Crashout ┃💰┃ MC : $488.5K ┃🎯┃ TARGET MC :
- $Stance looking healthy according to openclaw. adding more ┃🧫┃ $Stance ┃💰┃ MC : $15.9K ┃🎯┃ TARGET MC : $217.7K ┃📈┃
- openclaw on $HIMS —signals all pointing up ┃🧫┃ $HIMS ┃💰┃ MC : $213.6K ┃🎯┃ TARGET MC : $2.6M ┃📈┃ POTENTIAL : +1127%
- $AstroDOGE momentum score just came in hot on openclaw 🔥 ┃🧫┃ $AstroDOGE ┃💰┃ MC : $584.0K ┃🎯┃ TARGET MC : $7.6M ┃📈┃
- Thanks to Claude’s “shutdown” of Openclaw, I discovered @ollama. Couldn’t be happier with GLM 5.1 and Kimi 2.5. Now I do
- @iambensmart @OpenClawGeneva @tencentcloud @openclaw It was epic!
- @jukan05 Hi Jukan, thanks for sharing Fin’s take! As an AI Researcher, I have a slightly different observation on why CP
- @Js_Liim @Jez_Cryptoz @ritualfnd @ritualnet @joshsimenhoff @dunken9718 @cryptooflashh @0xMadScientist Testnet soon + age
- openclaw on $Stance —momentum strong, risk low. easy Was Made Using - @OpenClawAIX https://t.co/PqmJNJQult
- openclaw on $SOULANA —momentum strong, risk low. easy ┃🧫┃ $SOULANA ┃💰┃ MC : $280.5K ┃🎯┃ TARGET MC : $3.9M ┃📈┃ POTEN
- $Crashout momentum score just came in hot on openclaw 🔥 ┃🧫┃ $Crashout ┃💰┃ MC : $488.5K ┃🎯┃ TARGET MC : $6.9M ┃📈┃ PO
- $Crashout looking like a clean entry according to openclaw ┃🧫┃ $Crashout ┃💰┃ MC : $488.5K ┃🎯┃ TARGET MC : $5.4M ┃📈┃
- openclaw called it on $SOLANALIFE again. the model doesn't miss ┃🧫┃ $SOLANALIFE ┃💰┃ MC : $170.1K ┃🎯┃ TARGET MC : $2.
- $SOLANALIFE looking like a clean entry according to openclaw ┃🧫┃ $SOLANALIFE ┃💰┃ MC : $170.1K ┃🎯┃ TARGET MC : $2.1M
- $ASTEROID looking healthy according to openclaw. adding more ┃🧫┃ $ASTEROID ┃💰┃ MC : $2.4M ┃🎯┃ TARGET MC : $35.6M ┃📈
- @garrytan the mechanic tax is why most OpenClaw wins are the people already comfortable in infrastructure. others hit th
- @garrytan I’m using Hermes to maintain my openclaw
- Codex 也可以用 Custom Provider 的 API，不用擔心 Token 用量了，反正 OpenClaw 我都用不完 Token 了。 https://t.co/mGGtjH9evv
- @garrytan Majority of my "AI time" is spent in Claude Code and Codex. Am I falling behind by not prioritising an OpenCla
Releases
- OpenClaw repo stats for 2026-04-18 — 359682 stars, 73216 forks, 1781 contributors
- openclaw-cli 2026.4.15 on Homebrew — Your own personal AI assistant
Community
- Harden subagent completion delivery — Summary - persist a per-run delivery claim and user-safe delivery payload for subagent completions - route active-parent completion announces queue-first with fail-closed semantics instead of...
- fix(whatsapp): support silentErrorReplies parity — Summary - add channels.whatsapp.silentErrorReplies to the WhatsApp config schema/types and UI hints - suppress gateway-synthesized WhatsApp error payloads when that flag is enabled, matching Telegram...
- fix(config): refuse to restore from a polluted backup containing redacted secrets (#68423) — Summary Closes #68423. maybeRecoverSuspiciousConfigRead would silently restore from ~/.openclaw/openclaw.json.bak whenever the live config looked clobbered (update-channel-only root). It did not...
- [Bug]: URL-containing messages can be dropped after in-place upgrade due to stale link-understanding chunk import — Bug type Regression (worked before, now fails) Beta release blocker No Summary On an npm-global OpenClaw 2026.4.15 gateway upgraded in place on Ubuntu, Feishu text messages containing URLs failed...
- [Bug]: Plugin install/update can materialize peer openclaw into plugin node_modules, causing mixed-runtime crashes — Bug summary openclaw update triggers plugin update sync, but third-party plugin install/update can materialize peerDependencies.openclaw into plugin-local node_modules/openclaw. This creates a...
- Fix delivery-ready verification blockers — Summary This PR closes the remaining delivery-ready verification blockers for the current branch. It completes three founder-visible fixes: - suppress non-actionable historical task delivery warning...
- Port stability remediation from installed dist into source — Summary Ports the proven installed-dist stability remediation into OpenClaw source for v2026.4.14 so the fixes survive rebuilds and future updates. What Changed - Added canonical session-route...
- [Bug]: Model ID with provider prefix (google/gemma-4-26b-a4b) routes to wrong provider when using LM Studio — Bug type Behavior bug (incorrect output/state without crash) Beta release blocker No Summary Model ID containing provider prefix (e.g. google/gemma-4-26b-a4b) is incorrectly routed to the provider's...
- onboard (ollama): populate cloud-only model list from ollama.com/api/tags — Summary - Problem: openclaw onboard in Ollama cloud-only mode populated the provider's model list from a hardcoded, stale array (kimi-k2.5:cloud, minimax-m2.7:cloud, glm-5.1:cloud). The local-only...
- fix(matrix): isolate credentials runtime import boundary (refs #50477) — Summary - Problem: matrix-js-sdk triggers INEFFECTIVE_DYNAMIC_IMPORT esbuild warnings and a "multiple entrypoints" runtime error because the package is reachable through both static and dynamic...
- feat(tencent): add bundled Tencent Cloud provider plugin (Tokenhub + Token Plan)
- [Bug]: Display 'Context limit exceeded. I've reset our conversation to start fresh - please try again' after installed 2026.4.15 — Bug type Regression (worked before, now fails) Beta release blocker No Summary After upgraded to 2026.4.15 from 2026.4.8, the following error dispalyed no matter what I input: ⚠️ Context limit...
- fix(discord): honor explicit accountId token resolution in cron deliveries — Summary - prevents explicit non-default Discord accountId sends from silently inheriting the default/top-level token - returns source: none when the explicit account has no token instead of falling...
- fix(session): only persist user-selected auth profile overrides across /new and /reset — Summary /new and /reset currently persist providerOverride, modelOverride, and authProfileOverride onto the successor session entry unconditionally — including overrides that were never set by the...
- fix(mcp): dispose bundled MCP runtimes after isolated one-shot runs — Summary - Problem: isolated cron, isolated heartbeat, and one-shot follow-up runs can create bundle MCP runtimes and leave them alive after the one-shot run ends, which causes connection/process...
- feat(android/chat): filter cron-delivery noise and strip <final> wrapper — Summary - filters Android chat history entries that are cron delivery completion markers ("A cron job ... just completed successfully") so they do not render as user messages - strips top-level ......
- feat: loop mode, hyperfocus refactor, steer mid-turn, llama.cpp thinking — Summary - /loop mode: nuevo modo de operación autónoma continua para el agente - AgentMode refactor: renombrado IntensiveMode → AgentMode e intensive → hyperfocus en toda la base de código - Steer...
- CLI: `openclaw models set <alias>` does not resolve aliases; creates bogus model entries — Summary openclaw models set does not resolve the alias to its underlying model id. Instead, it registers a new literal model entry using / as the id and writes that bogus id as the primary model....
- WhatsApp: Add quoted reply support for outbound messages
- feat(gateway): accept audio/file content blocks in /v1/chat/completions — Summary The OpenAI-compatible /v1/chat/completions endpoint currently recognises only text / input_text / image_url content parts and silently drops input_audio, file, and video_url. This brings it...
- Add channels.whatsapp.silentErrorReplies (parity with Telegram) — Summary Telegram supports channels.telegram.silentErrorReplies: boolean which suppresses gateway-synthesized error replies to the user. WhatsApp has no equivalent config option. Requesting parity....
- security(media): anchor sanitizeMimeType regex to reject malformed input — Summary - Problem: sanitizeMimeType in src/media-understanding/apply.ts matches MIME types with an unanchored regex, so malformed input silently truncates to a valid-looking prefix. Example:...
- feat(feishu): allow custom accountId during Feishu channel onboarding — Problem When adding a second Feishu IM channel in the same OpenClaw instance, the accountId is always forced to DEFAULT_ACCOUNT_ID ("default"), making it impossible to register multiple independent...
- fix(openrouter): replace legacy base URL during model-config merge — Summary - Problem: persisted per-agent/provider config could preserve OpenRouter's legacy base URL (https://openrouter.ai/v1) during model-config merge instead of adopting the canonical...
- [Bug]: claude-opus-4-7 rejected as "Unknown model" on 2026.4.15 despite being in codebase and accepted by Anthropic API — Bug type Regression (worked before, now fails) Beta release blocker No Summary After upgrading from OpenClaw 2026.4.2 to 2026.4.15, configuring anthropic/claude-opus-4-7 as agent primary model causes...
- fix(whatsapp): add application-level health probe to inbound monitor — Problem The WhatsApp inbound monitor relies on the Baileys socket connection state, but a connected socket doesn't guarantee the application layer is healthy. The socket can remain open while the...
- fix(cron): preserve silent tool-result `NO_REPLY` on empty final turn — Fixes #68452. Summary This fixes a cron silent-success bug in embedded runner finalization. If an isolated cron agentTurn receives a successful tool result whose content is exact NO_REPLY, but the...
- fix(webchat): preserve image attachments for text-only models — Summary - Problem: When the primary model (e.g. GLM-4.7) does not support image input, image attachments from webchat/Control UI are silently dropped. Users receive no feedback and cannot analyze...
- OpenClaw sponsors (150 total) — 150 sponsors supporting OpenClaw
- refactor: extract readCanonicalStat and add structural guards for stat field migration (#1709) — Merged via squash. Prepared head SHA: e92817f66f44254f78a27df12022bf7a1b1af5b9 Co-authored-by: momothemage Co-authored-by: momothemage Reviewed-by: @momothemage
- fix(api): align inspect security snapshot with static scan moderation — Include static scan results in the skill version security snapshot so inspect/API responses reflect the same moderation-relevant signal already used elsewhere. Also add regression coverage for...
- feat: allow moderators to transfer skill publishers (#1663)
- feat: include stats in package detail API response — Expose package detail stats through the shared API contract and the app client. This lands the original package detail stats work and folds in the follow-up cleanup to keep the response shape sourced...
- docs: point morrow skill at agent-morrow namespace
- Keep Notes & PKM entries alphabetized
News
- peterjohannmedina/Medina-Qwen3.5-27B-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
- I Gave an AI a Body. Here’s What Happened. — A personal reckoning with OpenClaw — the open-source assistant that doesn’t just answer, it...
- How I Built a Personal AI Agent That Runs on My Google Cloud VM — And Powers My Entire Portfolio
- I Built an AI Agent That Texts Me on Telegram — And It Does My Job Search
- How I Built a Personal AI Workspace with OpenClaw and Reframed My Development Workflow — When I first came across OpenClaw, I was not looking for another AI tool. Like most developers, I...
- I Built an AI That Texts People Back as Me — and Nobody Noticed — This is a submission for the OpenClaw Challenge. What I Built I...
- I Built a Self-Healing PR Monitor With OpenClaw (And It Caught Its Own Bugs)
- I Stopped Treating AI Like Magic — and Built Something Useful with OpenClaw — Most people approach AI like it’s supposed to “do everything.” I tried something different: I gave...
- Hermes Agent Review: 95.6K Stars, Self-Improving AI Agent (April 2026) — Hermes Agent is Nous Research's open-source AI agent framework, released February 25, 2026. Seven...
- How I Run OpenClaw in Production: 3 Months of Lessons Running an Autonomous AI Agent
- Famous Souls: Drop-in Personalities for Your OpenClaw Agent — We just launched Famous Souls — a free, open-source collection of SOUL.md personality files for...
- OpenClaw in Action: How I Built a Multi-Agent Command Center with 5 Autonomous Claude Sessions
- Orquestacion multiagente con Openclaw dockerizada.
- I gave an AI agent the keys to my business for 30 days. Here's what actually happened. — Not a polished case study. What I actually experienced letting an AI agent run pieces of my operation — voice calls, odds monitoring, content, and lead gen.
- I Registered My AI Agent on a Freelance Marketplace — Here's What Actually Happened
- Inside OpenClaw: How Agentic Assistants Work — Over the last couple of weeks, I’ve been digging into OpenClaw from an engineering perspective. What...
- OpenClaw Monetization: 8 Proven Ways to Generate Income in 2026 — How to Make Money with OpenClaw OpenClaw isn't just a productivity tool — it's a platform...
- OpenCopilot (Continued) — IntroductionContinue reading on Medium »
- Building a Safer OpenClaw on a VM: Why This “Secure Stack” Repo Matters — There are plenty of ways to get OpenClaw running on a VM.Continue reading on Medium »
- Your own personal OpenClaw AI Agent for just 2$/month — What if your own AI agent could run on your server, automate tasks, and reply on WhatsApp — without expensive bills and subscription fees?Continue reading on The Tech Show »
- OpenClaw AI Agent Ecosystem: Enterprise Impact and VC Funding Signals — Executive SummaryContinue reading on Medium »
- Journey with Hermes: Day 5 — What it really looks like working with personal AI assistantsContinue reading on Medium »
- This is how I saved 50 % of the AI token cost! — Today, I want to talk about the “AI tokens” because it has become a form of digital money. More and more, frontier models like Anthropic…Continue reading on Medium »
- OpenClaw hospedado ou OpenClaw autohospedado: qual escolher? — Compare o OpenClaw hospedado em servidor próprio e o OpenClaw auto-hospedado em termos de velocidade de configuração, controle…Continue reading on Medium »
- Your $20/mo AI Wrapper is Dead: Why OpenClaw is Making 60% of SaaS Tools Obsolete — You’re probably dropping $20 a month for ChatGPT, another $20 for Claude, and then paying GitHubCopliot/ClaudeCode or Make just to move…Continue reading on Medium »
- Instantly Claw Review — The OpenClaw Platform That Works for You 24/7 with Zero Tech Headaches — A recent survey found that over 70% of freelancers waste 15+ hours per week just prompting and re-prompting AI tools instead of getting…Continue reading on Medium »
- NemoClaw for the Enterprise: A Zero-Trust Setup for OpenClaw (Part 1) — How to give your AI agent a safe house: full shell access, without becoming a liability to your security team.Continue reading on Medium »
- 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...
- 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: