OpenClaw Newsletter - 2026-06-28
OpenClaw Newsletter
Sunday, June 28, 2026Top Stories
- fix: project visible session transcripts — What Problem This Solves Visible WhatsApp/group session surfaces were showing internal delivery artifacts as if they were user-visible chat history. On...
- fix(agents): attach read tool images to replies — Summary Persist trusted read tool image content into the inbound media store when the tool result has image blocks but no MEDIA: path or details.path. Queue the saved media://inbound/... reference...
- fix(heartbeat): rewrite stream-error placeholder leak replies before delivery (#97357) — What Problem This Solves Fixes #97357 — heartbeat fallback can deliver repeated internal stream-error placeholders as user-visible messages. When the primary model fails before producing content...
- fix(irc): prevent ghost nick collisions on rejoin after network delay — Summary - When an IRC connection experiences a network delay and rejoins, the nick collision fallback always appends the same suffix _, causing a ghost nick collision with the still-connected...
- fix(zai): reject onboarding command auth profiles — Related: #97313 What Problem This Solves Fixes the Z.AI command-string auth-profile pollution path where users could end up with zai:default storing onboarding command text such as openclaw onboard...
- fix(llm): preserve structured tool result text across providers — Summary Fixes #96857 by preserving structured non-image tool result content as text during provider context conversion instead of dropping it and falling back to (see attached image) / empty-output...
- [AI] fix(plugins): recognize document-extractors as a capability kind… — Summary PluginCapabilityKind in inspect-shape.ts did not include "document-extractors", so plugins declaring contracts.documentExtractors (like the bundled document-extract plugin) showed...
- Fix/zalo bound api json response reads — What Problem This Solves All Zalo Bot API calls flow through callZaloApi(), which still parsed every success response with unbounded response.json(). On Node, response.json() buffers the entire body...
- fix(acp): return updatedInput instead of behavior: allow in canUseTool — Claude Code >= 2.1 tightened its PermissionResult schema to accept only: - { updatedInput: } (for allow) - { behavior: "deny", message: } (for deny) OpenClaw previously returned { behavior: "allow" }...
- fix(litellm): pass Anthropic cache control — Closes #37966 What Problem This Solves Fixes an issue where users routing Claude through LiteLLM would not get Anthropic prompt-cache markers even when they explicitly configured cacheRetention,...
- fix(code-mode): handle QuickJS interrupted error on VM dispose — Summary Two fixes in code-mode: 1. Handle QuickJS interrupted errors on VM dispose File: src/agents/code-mode.worker.ts vm.dispose() can throw isQuickJsInterruptedError when the VM was forcefully...
- Spinen/msteams multi account — Related: #71058 What Problem This Solves Resolves a problem where Microsoft Teams could only be configured as one bot identity per OpenClaw gateway. That made it difficult to run multiple...
- i18n: improve Indonesian Control UI labels — Summary Translates remaining English strings in ui/src/i18n/locales/id.ts for the Agents, Debug, and Exec Approval panels in the Control UI. Syncs matching .i18n artifacts. Changes -...
- Fix/discord bound probe getme json reads — What Problem This Solves Discord health probes call GET /users/@me and still parse the success body with unbounded res.json(). On Node, response.json() buffers the entire body before parsing, so a...
- fix(compact): make /compact command cancelable via abortEmbeddedAgentRun — What Problem This Solves Fixes an issue where /compact could not be canceled via /stop. The queued compaction task ran without a synthetic active embedded-run handle, so /stop could not find and...
- fix(skills): never wipe target skills dir on sync to prevent data los… — …s on partial load syncSkillsToWorkspace used to call fsp.rm(targetSkillsDir) before recopying every skill. Any transient failure inside loadWorkspaceSkillEntries (a filesystem hiccup, a config parse...
- fix(infra): wrap provider auth resolution in timeout for status --usage --json — Summary - Wraps resolveProviderAuths() in the existing withTimeout helper so auth resolution (OAuth token refresh, plugin auth hooks, secret exec resolvers) cannot hang indefinitely - Returns empty...
- feat(telegram): support mini app URL buttons — Summary - Reuse shared presentation URL buttons in Telegram inline keyboards. - Render direct-chat HTTPS presentation buttons as Telegram Mini App web_app buttons. - Document the Telegram HTTPS,...
- WORKING: All Microsoft Issues and PRs (refresh) — Microsoft Ecosystem Issues & PRs Tracker > Purpose: Living checklist for maintainers to track all open Microsoft-related issues and PRs (Teams, Windows, WSL, Azure, M365/SharePoint). > > How to use:...
- fix(cron): propagate cleanupCliLiveSessionOnRunEnd to isolated cron CLI branch — What Problem This Solves Isolated cron CLI workers (src/cron/isolated-agent/run-executor.ts at line ~306) call runCliAgent without passing cleanupCliLiveSessionOnRunEnd, so worker processes spawned...
- High host load & slow responses caused by stale openclaw worker process accumulation — Description On 2026.4.29 (and persisting into 2026.4.30-beta.1), response times degrade to 2-3+ minutes per turn due to stale openclaw worker processes accumulating on the host and driving load...
- fix(health-monitor): add reconnect grace for gateway reconnects — Repairs the existing #45712 branch rather than replacing it. Summary: - add shared reconnectGraceMs/lastDisconnectAt handling so running-but-recently-disconnected channels are not restarted before...
- fix(control-ui): share gateway token scope across loopback hosts — Repair the existing #43254 branch rather than replacing it. This keeps gateway tokens in sessionStorage, shares token scope only across true same-machine loopback aliases, and preserves protocol,...
- fix(agents): accept multi-line ANNOUNCE_SKIP on final line — Summary - Fix isAnnounceSkip() to also accept multi-line sub-agent output where ANNOUNCE_SKIP appears as the final line after a newline, not just as the sole content. - Add focused unit tests for...
- v2026.4.25 regression: memory-core breaks agent by using wrong API key for embeddings — Bug Description After upgrading from 2026.4.24 to 2026.4.25, the memory-core plugin breaks agent message processing by using the wrong API key for embeddings. What changed Before (2026.4.24): Agent...
- fix(exec): decode Windows command output with codepage-aware streaming — Summary - Fix Windows exec output mojibake for CJK locales by decoding captured command output with the active Windows code page instead of assuming UTF-8. - Preserve decoder state across...
- [Feature]: Expose machine-wide hook policies — Summary Add an admin-managed, system-wide hook discovery tier whose entries always load, run before user-controlled hooks, and cannot be disabled by _openclaw hooks disable_. Problem to solve...
- Feishu @larksuiteoapi/node-sdk axios timeout from gateway process (Alibaba Cloud ECS / Node.js 24) — Description When the OpenClaw Gateway starts with Feishu channel enabled in websocket connection mode, the @larksuiteoapi/node-sdk embedded axios calls to Feishu API (open.feishu.cn) consistently...
- OpenClaw repo stats for 2026-06-28 — 380783 stars, 79770 forks, 2680 contributors
- OpenClaw sponsors (111 total) — 111 sponsors supporting OpenClaw
- fix(security): tolerate transient scan claim failures — Treats transient partial claim failures as nonfatal for Codex security scan workers while keeping no-work claim outages fatal. Validation: targeted worker test, format check, git diff --check,...
- fix: harden publisher OG profile images
- fix: show temporal abuse signal pressure (#2920)
- fix: repair publisher abuse dashboard display (#2919)
- fix(ci): right-size security dataset snapshots — Right-size the ClawHub security dataset snapshot workflow after the scheduled run exposed a 120-minute shard timeout. Raises the default shard split while preserving hosted-runner max parallelism,...
- fix: reduce static scanner false positives — Reduce false positives in ClawHub static moderation scans while preserving dangerous child_process detection and secret literal coverage.
- Iter435/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...
- Google restricting Google AI Pro/Ultra subscribers for using OpenClaw
- Anthropic says OpenClaw-style Claude CLI usage is allowed again
- Ask HN: Who is using OpenClaw?
- Show HN: Klaus – OpenClaw on a VM, batteries included
- Show HN: DenchClaw – Local CRM on Top of OpenClaw
- My OpenClaw MCP Server Said 'OK' But Returned Nothing. I Built a 40-Line Health Check That Saved My Mornings. — My OpenClaw MCP Server Said "OK" But Returned Nothing. I Built a 40-Line Health Check That...
- 4 Safety Boundaries Your AI Agent Needs Before Production (And How to Wire Them) — Most AI agents go into production with zero guardrails. Here's the 4-boundary framework I run on every agent before it touches real data — and the exact code to implement each one.
- I Run a Self-Improvement Loop on My OpenClaw Agent Every Night. Here's What I Learned. — Last month my OpenClaw agent kept making the same mistake: it would run a health check, the script...
- Bukan Lagi Sekadar Chatbot: Catatan dari Diskusi OpenClaw — Ketika AI mulai mengingat dan bertindak, pertanyaannya bukan hanya seberapa pintar modelnya, tetapi juga seberapa aman aksesnya.Continue reading on Medium »
- Implementing Gbrain for a company — Part 1 — Why Gbrain (Pros and Cons)Continue reading on Medium »
- How We Solved OpenClaw’s Biggest Problem: Why We Moved from Local Ollama Models to MiniMax Cloud — When we first integrated OpenClaw into our workflow, we had a simple goal:Continue reading on Medium »
- HireFlow — Multi-Agent Hiring Pipeline With OpenClaw — I Built a Multi-Agent Pipeline as a POC — Here’s What I LearnedContinue reading on Medium »
- Sending Data to Template Files in Django — Complete Beginner’s Guide — IntroductionContinue reading on Medium »
- The AI Agent That Lives on Your Machine — You have heard the hype about AI agents. Every week there is a new announcement. Another framework. Another promise of automation. Most of…Continue reading on Medium »
- Parallel Free Search Lands in OpenClaw: A Zero-Key Web Search MCP, Tested Against Tavily — OpenClaw just shipped parallel-free — a no-account, no-API-key web search provider built on Parallel's MCP.Continue reading on Level Up Coding »
- my openclaw analyst flagged $three as a buy. noted 🦞 $three — quick check • Current MC: $3.2M • 24h Volume: $1.0M •
- been watching $MANIFEST through openclaw, momentum is building $MANIFEST · health 🟩🟩🟩⬜⬜⬜⬜⬜⬜⬜ mc $19.0M · vol $1.7M ·
- $CHILLGUY looking like a clean entry according to openclaw $CHILLGUY MC $8.8M → $34.9M (+299%) 24h vol $133.0K Was
- my openclaw scan on $WOULD just confirmed what i was thinking $WOULD | MC $81.7M → $241.8M | Vol $14.6K | +196% Was
- openclaw flagged $jellyjelly as a strong hold. not surprised $jellyjelly | MC $55.8M → $149.0M | Vol $385.5K | +167%
- just ran $neet through openclaw. risk is low, upside is real $neet MC $22.9M → $77.3M (+237%) 24h vol $370.8K Was
- openclaw called $arc bullish. i'm listening [SCAN] $arc [MC] $87.2M [VOL] $1.4M [LIQ] $4.8M [TGT] $258.9M ×3
- my openclaw analyst flagged $swarms as a buy. noted 🦞 $swarms ┃ mc $6.3M ┃ vol $99.5K ┃ liq $1.1M ┃ tgt $24.5M ┃ +291%
- just ran $Coinomi through openclaw. risk is low, upside is real $Coinomi | MC $992.6K → $3.8M | Vol $32.3K | +283% W
- openclaw been right on $KINS twice already. trusting the scan > KINS MKT CAP $13.8M 24H VOL $3.1M LIQU
- openclaw flagged $PYTHIA as a strong hold. not surprised $PYTHIA | MC $16.9M → $61.3M | Vol $256.4K | +263% Was Made
- just checked $VINE on openclaw—forecast looking clean 🦞 $VINE · health 🟥⬜⬜⬜⬜⬜⬜⬜⬜⬜ mc $10.4M · vol $84.9K · liq $1.5M
- @GergelyOrosz @karpathy People have been able to set this up with custom tools using openclaw/hermes for months. Their i
- just ran $Fartcoin through openclaw. risk is low, upside is real $Fartcoin | MC $130.6M → $355.3M | Vol $3.3M | +172%
- @markproduct Still using it in pockets. Codex/OpenClaw for agents, Cursor when I want to stay closer to the code.
- RT @elonmusk: People giving OpenClaw root access to their entire life https://t.co/jqW0FwiW03
- RT @JulianGoldieSEO: There's a OpenClaw FULL COURSE 5 HOURS (Build & Automate Anything) https://t.co/7QteSi8RGw
- RT @AYi_AInotes: 现在用AI做视频可以跟喝水一样简单，不需要再付个700多块的剪映SVIP， 装这6个2026 年最顶的插件和skills就够了， 链接直接丢给你的AI Agent（Claude Code、Cursor、
- Openclaw + whoop mcp to monitor my health!!! https://t.co/OYjRWIlOhT
- $$NRG looking healthy according to openclaw. adding more $$NRG ▁▂▃▅▇ → $9.8M mc $2.7M · vol $13.9K · liq $177.0K · +2
Trending on X
- my openclaw analyst flagged $three as a buy. noted 🦞 $three — quick check • Current MC: $3.2M • 24h Volume: $1.0M •
- been watching $MANIFEST through openclaw, momentum is building $MANIFEST · health 🟩🟩🟩⬜⬜⬜⬜⬜⬜⬜ mc $19.0M · vol $1.7M ·
- $CHILLGUY looking like a clean entry according to openclaw $CHILLGUY MC $8.8M → $34.9M (+299%) 24h vol $133.0K Was
- my openclaw scan on $WOULD just confirmed what i was thinking $WOULD | MC $81.7M → $241.8M | Vol $14.6K | +196% Was
- openclaw flagged $jellyjelly as a strong hold. not surprised $jellyjelly | MC $55.8M → $149.0M | Vol $385.5K | +167%
- just ran $neet through openclaw. risk is low, upside is real $neet MC $22.9M → $77.3M (+237%) 24h vol $370.8K Was
- openclaw called $arc bullish. i'm listening [SCAN] $arc [MC] $87.2M [VOL] $1.4M [LIQ] $4.8M [TGT] $258.9M ×3
- my openclaw analyst flagged $swarms as a buy. noted 🦞 $swarms ┃ mc $6.3M ┃ vol $99.5K ┃ liq $1.1M ┃ tgt $24.5M ┃ +291%
- just ran $Coinomi through openclaw. risk is low, upside is real $Coinomi | MC $992.6K → $3.8M | Vol $32.3K | +283% W
- openclaw been right on $KINS twice already. trusting the scan > KINS MKT CAP $13.8M 24H VOL $3.1M LIQU
- openclaw flagged $PYTHIA as a strong hold. not surprised $PYTHIA | MC $16.9M → $61.3M | Vol $256.4K | +263% Was Made
- just checked $VINE on openclaw—forecast looking clean 🦞 $VINE · health 🟥⬜⬜⬜⬜⬜⬜⬜⬜⬜ mc $10.4M · vol $84.9K · liq $1.5M
- @GergelyOrosz @karpathy People have been able to set this up with custom tools using openclaw/hermes for months. Their i
- just ran $Fartcoin through openclaw. risk is low, upside is real $Fartcoin | MC $130.6M → $355.3M | Vol $3.3M | +172%
- @markproduct Still using it in pockets. Codex/OpenClaw for agents, Cursor when I want to stay closer to the code.
- RT @elonmusk: People giving OpenClaw root access to their entire life https://t.co/jqW0FwiW03
- RT @JulianGoldieSEO: There's a OpenClaw FULL COURSE 5 HOURS (Build & Automate Anything) https://t.co/7QteSi8RGw
- RT @AYi_AInotes: 现在用AI做视频可以跟喝水一样简单，不需要再付个700多块的剪映SVIP， 装这6个2026 年最顶的插件和skills就够了， 链接直接丢给你的AI Agent（Claude Code、Cursor、
- Openclaw + whoop mcp to monitor my health!!! https://t.co/OYjRWIlOhT
- $$NRG looking healthy according to openclaw. adding more $$NRG ▁▂▃▅▇ → $9.8M mc $2.7M · vol $13.9K · liq $177.0K · +2
Releases
- OpenClaw repo stats for 2026-06-28 — 380783 stars, 79770 forks, 2680 contributors
Community
- fix: project visible session transcripts — What Problem This Solves Visible WhatsApp/group session surfaces were showing internal delivery artifacts as if they were user-visible chat history. On...
- fix(agents): attach read tool images to replies — Summary Persist trusted read tool image content into the inbound media store when the tool result has image blocks but no MEDIA: path or details.path. Queue the saved media://inbound/... reference...
- fix(heartbeat): rewrite stream-error placeholder leak replies before delivery (#97357) — What Problem This Solves Fixes #97357 — heartbeat fallback can deliver repeated internal stream-error placeholders as user-visible messages. When the primary model fails before producing content...
- fix(irc): prevent ghost nick collisions on rejoin after network delay — Summary - When an IRC connection experiences a network delay and rejoins, the nick collision fallback always appends the same suffix _, causing a ghost nick collision with the still-connected...
- fix(zai): reject onboarding command auth profiles — Related: #97313 What Problem This Solves Fixes the Z.AI command-string auth-profile pollution path where users could end up with zai:default storing onboarding command text such as openclaw onboard...
- fix(llm): preserve structured tool result text across providers — Summary Fixes #96857 by preserving structured non-image tool result content as text during provider context conversion instead of dropping it and falling back to (see attached image) / empty-output...
- [AI] fix(plugins): recognize document-extractors as a capability kind… — Summary PluginCapabilityKind in inspect-shape.ts did not include "document-extractors", so plugins declaring contracts.documentExtractors (like the bundled document-extract plugin) showed...
- Fix/zalo bound api json response reads — What Problem This Solves All Zalo Bot API calls flow through callZaloApi(), which still parsed every success response with unbounded response.json(). On Node, response.json() buffers the entire body...
- fix(acp): return updatedInput instead of behavior: allow in canUseTool — Claude Code >= 2.1 tightened its PermissionResult schema to accept only: - { updatedInput: } (for allow) - { behavior: "deny", message: } (for deny) OpenClaw previously returned { behavior: "allow" }...
- fix(litellm): pass Anthropic cache control — Closes #37966 What Problem This Solves Fixes an issue where users routing Claude through LiteLLM would not get Anthropic prompt-cache markers even when they explicitly configured cacheRetention,...
- fix(code-mode): handle QuickJS interrupted error on VM dispose — Summary Two fixes in code-mode: 1. Handle QuickJS interrupted errors on VM dispose File: src/agents/code-mode.worker.ts vm.dispose() can throw isQuickJsInterruptedError when the VM was forcefully...
- Spinen/msteams multi account — Related: #71058 What Problem This Solves Resolves a problem where Microsoft Teams could only be configured as one bot identity per OpenClaw gateway. That made it difficult to run multiple...
- i18n: improve Indonesian Control UI labels — Summary Translates remaining English strings in ui/src/i18n/locales/id.ts for the Agents, Debug, and Exec Approval panels in the Control UI. Syncs matching .i18n artifacts. Changes -...
- Fix/discord bound probe getme json reads — What Problem This Solves Discord health probes call GET /users/@me and still parse the success body with unbounded res.json(). On Node, response.json() buffers the entire body before parsing, so a...
- fix(compact): make /compact command cancelable via abortEmbeddedAgentRun — What Problem This Solves Fixes an issue where /compact could not be canceled via /stop. The queued compaction task ran without a synthetic active embedded-run handle, so /stop could not find and...
- fix(skills): never wipe target skills dir on sync to prevent data los… — …s on partial load syncSkillsToWorkspace used to call fsp.rm(targetSkillsDir) before recopying every skill. Any transient failure inside loadWorkspaceSkillEntries (a filesystem hiccup, a config parse...
- fix(infra): wrap provider auth resolution in timeout for status --usage --json — Summary - Wraps resolveProviderAuths() in the existing withTimeout helper so auth resolution (OAuth token refresh, plugin auth hooks, secret exec resolvers) cannot hang indefinitely - Returns empty...
- feat(telegram): support mini app URL buttons — Summary - Reuse shared presentation URL buttons in Telegram inline keyboards. - Render direct-chat HTTPS presentation buttons as Telegram Mini App web_app buttons. - Document the Telegram HTTPS,...
- WORKING: All Microsoft Issues and PRs (refresh) — Microsoft Ecosystem Issues & PRs Tracker > Purpose: Living checklist for maintainers to track all open Microsoft-related issues and PRs (Teams, Windows, WSL, Azure, M365/SharePoint). > > How to use:...
- fix(cron): propagate cleanupCliLiveSessionOnRunEnd to isolated cron CLI branch — What Problem This Solves Isolated cron CLI workers (src/cron/isolated-agent/run-executor.ts at line ~306) call runCliAgent without passing cleanupCliLiveSessionOnRunEnd, so worker processes spawned...
- High host load & slow responses caused by stale openclaw worker process accumulation — Description On 2026.4.29 (and persisting into 2026.4.30-beta.1), response times degrade to 2-3+ minutes per turn due to stale openclaw worker processes accumulating on the host and driving load...
- fix(health-monitor): add reconnect grace for gateway reconnects — Repairs the existing #45712 branch rather than replacing it. Summary: - add shared reconnectGraceMs/lastDisconnectAt handling so running-but-recently-disconnected channels are not restarted before...
- fix(control-ui): share gateway token scope across loopback hosts — Repair the existing #43254 branch rather than replacing it. This keeps gateway tokens in sessionStorage, shares token scope only across true same-machine loopback aliases, and preserves protocol,...
- fix(agents): accept multi-line ANNOUNCE_SKIP on final line — Summary - Fix isAnnounceSkip() to also accept multi-line sub-agent output where ANNOUNCE_SKIP appears as the final line after a newline, not just as the sole content. - Add focused unit tests for...
- v2026.4.25 regression: memory-core breaks agent by using wrong API key for embeddings — Bug Description After upgrading from 2026.4.24 to 2026.4.25, the memory-core plugin breaks agent message processing by using the wrong API key for embeddings. What changed Before (2026.4.24): Agent...
- fix(exec): decode Windows command output with codepage-aware streaming — Summary - Fix Windows exec output mojibake for CJK locales by decoding captured command output with the active Windows code page instead of assuming UTF-8. - Preserve decoder state across...
- [Feature]: Expose machine-wide hook policies — Summary Add an admin-managed, system-wide hook discovery tier whose entries always load, run before user-controlled hooks, and cannot be disabled by _openclaw hooks disable_. Problem to solve...
- Feishu @larksuiteoapi/node-sdk axios timeout from gateway process (Alibaba Cloud ECS / Node.js 24) — Description When the OpenClaw Gateway starts with Feishu channel enabled in websocket connection mode, the @larksuiteoapi/node-sdk embedded axios calls to Feishu API (open.feishu.cn) consistently...
- OpenClaw sponsors (111 total) — 111 sponsors supporting OpenClaw
- fix(security): tolerate transient scan claim failures — Treats transient partial claim failures as nonfatal for Codex security scan workers while keeping no-work claim outages fatal. Validation: targeted worker test, format check, git diff --check,...
- fix: harden publisher OG profile images
- fix: show temporal abuse signal pressure (#2920)
- fix: repair publisher abuse dashboard display (#2919)
- fix(ci): right-size security dataset snapshots — Right-size the ClawHub security dataset snapshot workflow after the scheduled run exposed a 120-minute shard timeout. Raises the default shard split while preserving hosted-runner max parallelism,...
- fix: reduce static scanner false positives — Reduce false positives in ClawHub static moderation scans while preserving dangerous child_process detection and secret literal coverage.
News
- Iter435/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...
- Google restricting Google AI Pro/Ultra subscribers for using OpenClaw
- Anthropic says OpenClaw-style Claude CLI usage is allowed again
- Ask HN: Who is using OpenClaw?
- Show HN: Klaus – OpenClaw on a VM, batteries included
- Show HN: DenchClaw – Local CRM on Top of OpenClaw
- My OpenClaw MCP Server Said 'OK' But Returned Nothing. I Built a 40-Line Health Check That Saved My Mornings. — My OpenClaw MCP Server Said "OK" But Returned Nothing. I Built a 40-Line Health Check That...
- 4 Safety Boundaries Your AI Agent Needs Before Production (And How to Wire Them) — Most AI agents go into production with zero guardrails. Here's the 4-boundary framework I run on every agent before it touches real data — and the exact code to implement each one.
- I Run a Self-Improvement Loop on My OpenClaw Agent Every Night. Here's What I Learned. — Last month my OpenClaw agent kept making the same mistake: it would run a health check, the script...
- Bukan Lagi Sekadar Chatbot: Catatan dari Diskusi OpenClaw — Ketika AI mulai mengingat dan bertindak, pertanyaannya bukan hanya seberapa pintar modelnya, tetapi juga seberapa aman aksesnya.Continue reading on Medium »
- Implementing Gbrain for a company — Part 1 — Why Gbrain (Pros and Cons)Continue reading on Medium »
- How We Solved OpenClaw’s Biggest Problem: Why We Moved from Local Ollama Models to MiniMax Cloud — When we first integrated OpenClaw into our workflow, we had a simple goal:Continue reading on Medium »
- HireFlow — Multi-Agent Hiring Pipeline With OpenClaw — I Built a Multi-Agent Pipeline as a POC — Here’s What I LearnedContinue reading on Medium »
- Sending Data to Template Files in Django — Complete Beginner’s Guide — IntroductionContinue reading on Medium »
- The AI Agent That Lives on Your Machine — You have heard the hype about AI agents. Every week there is a new announcement. Another framework. Another promise of automation. Most of…Continue reading on Medium »
- Parallel Free Search Lands in OpenClaw: A Zero-Key Web Search MCP, Tested Against Tavily — OpenClaw just shipped parallel-free — a no-account, no-API-key web search provider built on Parallel's MCP.Continue reading on Level Up Coding »
Don't miss what's next. Subscribe to Openclaw Newsletter: