[{"data":1,"prerenderedAt":1812},["ShallowReactive",2],{"blog-post-hermes-response-truncated-fix":3,"related-posts-hermes-response-truncated-fix":462},{"id":4,"title":5,"author":6,"body":10,"category":439,"date":440,"description":441,"extension":442,"featured":443,"image":444,"imageHeight":445,"imageWidth":445,"meta":446,"navigation":447,"path":448,"readingTime":449,"seo":450,"seoTitle":451,"stem":452,"tags":453,"updatedDate":440,"__hash__":461},"blog/blog/hermes-response-truncated-fix.md","Hermes Agent \"Response Truncated Due to Output Length Limit\": 5 Causes and Fixes",{"name":7,"role":8,"avatar":9},"Shabnam Katoch","Growth Head","/img/avatars/shabnam-profile.jpeg",{"type":11,"value":12,"toc":422},"minimark",[13,17,20,32,35,38,43,50,57,70,87,105,118,122,128,134,144,154,163,167,170,173,179,196,202,206,212,218,228,246,249,253,259,265,274,279,283,289,300,310,322,333,336,346,350,355,361,365,387,391,405,409,415,419],[14,15,16],"p",{},"The agent starts generating. Mid-sentence, it stops. \"Response truncated due to output length limit.\" The output is useless. The conversation is broken. Here are five causes from real GitHub issues and the fix for each.",[14,18,19],{},"A Chinese user posted on X during Labor Day weekend: \"My Hermes keeps throwing 'Response truncated due to output length limit.' I've given up on it. Let it starve.\"",[14,21,22,23,27,28,31],{},"That's the vibe. The error is maddening because it looks like a simple limit you should be able to increase. But ",[24,25,26],"code",{},"max_tokens"," in ",[24,29,30],{},"config.yaml"," had no effect until recently. The compression system has a math bug that prevents it from firing. And the hardcoded output limits can drain your OpenRouter credits before the model generates a single token.",[14,33,34],{},"GitHub issue #7237 documents the core complaint: \"This truncates the output mid-stream, breaks the conversation flow, and prevents users from receiving complete, usable answers.\"",[14,36,37],{},"Here are five causes, ranked by how often they're the actual problem.",[39,40,42],"h2",{"id":41},"cause-1-max_tokens-from-configyaml-never-reaches-the-api-confirmed-bug","Cause 1: max_tokens from config.yaml never reaches the API (confirmed bug)",[14,44,45],{},[46,47],"img",{"alt":48,"src":49},"Hermes max_tokens config bug: value set in config.yaml but never reaches API request","/img/blog/hermes-response-truncated-fix-config-bug.jpg",[14,51,52,56],{},[53,54,55],"strong",{},"GitHub issue #4404:"," \"model.max_tokens in config.yaml has no effect. The setting is never passed to AIAgent.\"",[14,58,59,62,63,66,67,69],{},[53,60,61],{},"What happens:"," You set ",[24,64,65],{},"model.max_tokens: 8192"," in your config. The agent ignores it. The API request goes out without ",[24,68,26],{},". The provider uses its default (often 2,048 or 4,096). Your response gets truncated at a limit you didn't set.",[14,71,72,75,76,79,80,83,84,86],{},[53,73,74],{},"The bug confirmed:"," A developer patched the code to log what ",[24,77,78],{},"_build_api_kwargs()"," actually sends. Result: ",[24,81,82],{},"self.max_tokens=None",". The config value exists but the code path from ",[24,85,30],{}," → AIAgent → API request is broken.",[14,88,89,92,93,96,97,100,101,104],{},[53,90,91],{},"The fix:"," A community PR exists on ",[24,94,95],{},"fix/model-max-tokens-config"," branch. If you're on v0.13.0+, check if the fix is merged. If not, the workaround: set ",[24,98,99],{},"HERMES_MAX_TOKENS=8192"," as an environment variable in ",[24,102,103],{},"~/.hermes/.env",". The env var path works even when the config path doesn't.",[14,106,107,110,111,114,115,117],{},[53,108,109],{},"The frustrating truth:"," The most common cause of truncation is a config value that looks correct but is silently ignored. Always verify with ",[24,112,113],{},"hermes chat -q \"write a 500-word essay\" --verbose"," and check the API request logs for the actual ",[24,116,26],{}," value sent.",[39,119,121],{"id":120},"cause-2-provider-default-output-limit-is-too-low-especially-ollama","Cause 2: Provider default output limit is too low (especially Ollama)",[14,123,124,127],{},[53,125,126],{},"What happens with Ollama:"," The default context window is 2,048 tokens. Not 8,192. Not the model's maximum. Two thousand and forty-eight. A 500-word response easily exceeds this.",[14,129,130,133],{},[53,131,132],{},"The fix for Ollama:"," Create a Modelfile that sets the correct limits:",[135,136,141],"pre",{"className":137,"code":139,"language":140},[138],"language-text","FROM hermes3:8b\nPARAMETER num_ctx 8192\nPARAMETER num_predict 1024\n","text",[24,142,139],{"__ignoreMap":143},"",[14,145,146,149,150,153],{},[24,147,148],{},"num_ctx"," is the total context window. ",[24,151,152],{},"num_predict"," is the max output tokens. Without these, Ollama defaults to 2,048 total, leaving almost no room for output after the system prompt and conversation history consume their share.",[14,155,156,157,162],{},"For the complete model configuration guide, our ",[158,159,161],"a",{"href":160},"/blog/openclaw-best-practices","best practices post"," covers per-model output limit configuration.",[39,164,166],{"id":165},"cause-3-context-window-full-no-room-left-for-output","Cause 3: Context window full (no room left for output)",[14,168,169],{},"Here's where most people get it wrong.",[14,171,172],{},"The context window is shared between input and output. If your model has a 64K context window and your conversation history + system prompt + tool definitions consume 60K tokens, only 4K tokens remain for the response. The model starts generating, hits the ceiling at 4K, and truncates.",[14,174,175],{},[46,176],{"alt":177,"src":178},"Context window math: input plus output share the same budget, leaving little room when conversation history is large","/img/blog/hermes-response-truncated-fix-context-math.jpg",[14,180,181,183,184,187,188,191,192,195],{},[53,182,91],{}," Run ",[24,185,186],{},"/usage"," in your chat to see current context consumption. If you're above 70%, run ",[24,189,190],{},"/compress"," to summarize the conversation history and free space. Or start a new session with ",[24,193,194],{},"hermes chat --new",".",[14,197,198,199,201],{},"The official FAQ confirms: \"Use ",[24,200,190],{}," regularly during long sessions. It summarizes the conversation history and reduces token usage significantly while preserving context.\"",[39,203,205],{"id":204},"cause-4-compression-never-triggers-the-math-bug","Cause 4: Compression never triggers (the math bug)",[14,207,208,211],{},[53,209,210],{},"GitHub issue #14690 (P1, open):"," \"Context auto-compression never triggers when context_length equals MINIMUM_CONTEXT_LENGTH (64000).\"",[14,213,214,217],{},[53,215,216],{},"The bug:"," The compression threshold calculation uses 64,000 as an absolute floor. If your model's context is 64K (common for local models with parallel slots), the threshold becomes 100% of the context window. Compression can't trigger because the API errors out before the threshold is reached.",[14,219,220,223,224,227],{},[53,221,222],{},"The math:"," ",[24,225,226],{},"threshold_tokens = max(int(64000 * 0.7), 64000) = max(44800, 64000) = 64000",". Threshold = 100%. Compression never fires. Context grows until the API rejects the request. Response gets truncated.",[14,229,230,232,233,27,236,238,239,242,243,245],{},[53,231,91],{}," Set ",[24,234,235],{},"model.context_length",[24,237,30],{}," to a value above 64,000 (e.g., ",[24,240,241],{},"128000",") so the threshold calculation produces a meaningful percentage. Or manually run ",[24,244,190],{}," before the context fills.",[14,247,248],{},"If debugging config values that are silently ignored, Ollama defaults that nobody documents, context window math bugs, and compression thresholds that never trigger sounds like more framework internals than agent building, BetterClaw's smart context management handles all of this at the platform level. No max_tokens configuration. No compression commands. No context math. The platform manages output limits and context automatically. Free tier with 1 agent and BYOK. $19/month per agent for Pro.",[39,250,252],{"id":251},"cause-5-hardcoded-max_tokens-reserves-too-many-credits-on-openrouter","Cause 5: Hardcoded max_tokens reserves too many credits on OpenRouter",[14,254,255],{},[46,256],{"alt":257,"src":258},"OpenRouter credit reservation: Hermes requests max_tokens=64000, OpenRouter reserves full amount as collateral, balance insufficient","/img/blog/hermes-response-truncated-fix-openrouter-credits.jpg",[14,260,261,264],{},[53,262,263],{},"GitHub issue #22879:"," \"Hermes hardcodes max_tokens to each model's maximum output (e.g., 64000 for Claude Sonnet/Haiku 4.5). OpenRouter reserves the full max_tokens as collateral before allowing the call.\"",[14,266,267,269,270,273],{},[53,268,61],{}," You have $10 in OpenRouter credits. Hermes requests ",[24,271,272],{},"max_tokens=64000",". OpenRouter reserves $10+ worth of credits upfront. Your balance can't cover the reservation. OpenRouter returns HTTP 402 (\"requires more credits\"). The actual response would be 50-500 tokens. The reservation is 64,000.",[14,275,276,278],{},[53,277,91],{}," This is tagged as a feature request (#22879). Until it's resolved, add more credits to your OpenRouter account (enough to cover the worst-case reservation), or switch to a direct provider (Anthropic direct doesn't pre-reserve credits).",[39,280,282],{"id":281},"the-diagnostic-checklist","The diagnostic checklist",[14,284,285],{},[46,286],{"alt":287,"src":288},"Four-step Hermes truncation diagnostic checklist: verbose log check, usage command, Ollama Modelfile, context_length config","/img/blog/hermes-response-truncated-fix-checklist.jpg",[14,290,291,223,294,296,297,299],{},[53,292,293],{},"Step 1:",[24,295,113],{}," ... Check the API request for actual ",[24,298,26],{}," value.",[14,301,302,223,305,307,308,195],{},[53,303,304],{},"Step 2:",[24,306,186],{}," in an active chat ... How much context is consumed? If above 70%, run ",[24,309,190],{},[14,311,312,315,316,318,319,321],{},[53,313,314],{},"Step 3:"," Check your Ollama Modelfile ... Is ",[24,317,148],{}," set? Is ",[24,320,152],{}," set? Defaults are too low.",[14,323,324,327,328,27,330,332],{},[53,325,326],{},"Step 4:"," Check ",[24,329,235],{},[24,331,30],{}," ... Is it 64,000? If so, the compression bug (#14690) applies.",[14,334,335],{},"The truncation error is Hermes's way of saying \"the model ran out of room.\" But \"ran out of room\" has five different causes, and the error message doesn't tell you which one. Config values that are ignored. Provider defaults that are undocumented. Context that fills silently. Compression that can't fire. Credit reservations that block the call. Same error. Five different problems.",[14,337,338,339,345],{},"If you want an agent where context management is handled automatically and you never see \"response truncated,\" ",[158,340,344],{"href":341,"rel":342},"https://app.betterclaw.io/sign-in",[343],"nofollow","give BetterClaw a try",". Free tier with 1 agent and BYOK. $19/month per agent for Pro. Smart context management. No truncation. No compression commands. The agent speaks. The response completes.",[39,347,349],{"id":348},"frequently-asked-questions","Frequently Asked Questions",[351,352,354],"h3",{"id":353},"what-does-response-truncated-due-to-output-length-limit-mean-in-hermes","What does \"Response truncated due to output length limit\" mean in Hermes?",[14,356,357,358,360],{},"It means the model's response was cut off before completion. The five causes: ",[24,359,26],{}," config not being sent to the API (bug #4404), provider default output limit too low (Ollama defaults to 2,048), context window full with no room for output, compression math bug preventing auto-compression (#14690), or OpenRouter credit reservation blocking the call (#22879).",[351,362,364],{"id":363},"how-do-i-increase-the-output-length-in-hermes-agent","How do I increase the output length in Hermes Agent?",[14,366,367,368,27,370,372,373,376,377,380,381,383,384,386],{},"Set ",[24,369,99],{},[24,371,103],{}," (the config.yaml path may not work due to bug #4404). For Ollama, create a Modelfile with ",[24,374,375],{},"PARAMETER num_ctx 8192"," and ",[24,378,379],{},"PARAMETER num_predict 1024",". For cloud providers, verify the ",[24,382,26],{}," value in verbose logs. Use ",[24,385,190],{}," regularly to free context space during long sessions.",[351,388,390],{"id":389},"why-does-compress-not-work-in-hermes","Why does /compress not work in Hermes?",[14,392,393,394,397,398,27,400,238,402,404],{},"If your model's ",[24,395,396],{},"context_length"," equals 64,000 (the MINIMUM_CONTEXT_LENGTH constant), auto-compression never triggers due to a math bug (#14690, P1, open). The threshold calculation produces 100% of the context window, which is unreachable. Fix: set ",[24,399,235],{},[24,401,30],{},[24,403,241],{},") so the threshold calculation works correctly.",[351,406,408],{"id":407},"why-does-hermes-use-all-my-openrouter-credits-on-one-message","Why does Hermes use all my OpenRouter credits on one message?",[14,410,411,412,414],{},"Hermes hardcodes ",[24,413,26],{}," to each model's maximum output (64,000 for Claude models). OpenRouter pre-reserves the full amount as credit collateral before allowing the call. Your $10 balance can't cover a 64K-token reservation even though the actual response would be 50 tokens. Fix: add more credits, or switch to a direct provider that doesn't pre-reserve. Issue #22879 tracks making this configurable.",[351,416,418],{"id":417},"does-betterclaw-have-the-same-truncation-problems","Does BetterClaw have the same truncation problems?",[14,420,421],{},"No. BetterClaw's smart context management handles output limits, context compression, and provider-specific configurations at the platform level. There's no max_tokens to configure, no compression threshold bug, and no credit reservation mismatch. The platform manages the context window so responses complete without truncation. Free tier with 1 agent and BYOK. $19/month per agent for Pro.",{"title":143,"searchDepth":423,"depth":423,"links":424},2,[425,426,427,428,429,430,431],{"id":41,"depth":423,"text":42},{"id":120,"depth":423,"text":121},{"id":165,"depth":423,"text":166},{"id":204,"depth":423,"text":205},{"id":251,"depth":423,"text":252},{"id":281,"depth":423,"text":282},{"id":348,"depth":423,"text":349,"children":432},[433,435,436,437,438],{"id":353,"depth":434,"text":354},3,{"id":363,"depth":434,"text":364},{"id":389,"depth":434,"text":390},{"id":407,"depth":434,"text":408},{"id":417,"depth":434,"text":418},"Troubleshooting","2026-05-19","Hermes \"Response truncated due to output length limit\" has 5 causes. Config bug, Ollama defaults, full context, compression math, OpenRouter credits. Fixes here.","md",false,"/img/blog/hermes-response-truncated-fix.jpg",null,{},true,"/blog/hermes-response-truncated-fix","9 min read",{"title":5,"description":441},"Hermes \"Response Truncated\": 5 Causes Fixed (2026)","blog/hermes-response-truncated-fix",[454,455,456,457,458,459,460],"Hermes response truncated","Hermes output length limit","Hermes Agent truncated fix","Hermes max_tokens","Hermes context window","Hermes compression bug","Hermes OpenRouter credits","9bJbGY_pISq93_jHnNmziIrOD9ipRxMD3d6ux8djbtY",[463,964,1363],{"id":464,"title":465,"author":466,"body":467,"category":439,"date":947,"description":948,"extension":442,"featured":443,"image":949,"imageHeight":445,"imageWidth":445,"meta":950,"navigation":447,"path":951,"readingTime":952,"seo":953,"seoTitle":954,"stem":955,"tags":956,"updatedDate":440,"__hash__":963},"blog/blog/claude-cowork-not-working-windows.md","Claude Cowork Not Working on Windows? Every Known Bug and the Best Workaround in 2026",{"name":7,"role":8,"avatar":9},{"type":11,"value":468,"toc":922},[469,474,477,481,488,491,505,512,521,528,532,547,551,554,558,561,565,568,572,575,579,582,586,589,595,599,602,605,608,611,614,623,627,630,633,636,643,650,654,665,668,676,680,683,686,689,692,695,698,704,708,712,715,741,744,748,755,778,781,785,788,792,795,810,816,820,823,831,835,838,850,852,857,860,865,879,884,894,899,910,915,918],[14,470,471],{},[53,472,473],{},"Claude Cowork fails on Windows for five reasons: (1) the CoworkVMService stops after reboot or sleep, (2) the \"yukonSilver\" platform detection bug marks capable systems as unsupported, (3) Windows Home edition lacks the full Hyper-V stack Cowork needs, (4) network conflicts with VPNs or Docker on the 172.16.0.0/24 range, and (5) corrupted installs from the old Squirrel installer. Each has a different fix.",[14,475,476],{},"Cowork shipped on Windows on February 10, 2026, and went GA across all paying subscribers on April 9, 2026. The Claude Code GitHub repo has been collecting Windows-specific bugs since launch: cryptic \"yukonSilver not supported\" errors, missing Cowork tabs on fully capable machines, and a VM service that resists removal. We've tracked the major failure modes and what actually fixes each one. No fluff.",[39,478,480],{"id":479},"try-this-first-restart-coworkvmservice","Try this first: restart CoworkVMService",[14,482,483,484,487],{},"Before anything else, check whether Cowork's background service is actually running. CoworkVMService ships with startup type ",[53,485,486],{},"Manual",", which means it stops after reboots, Windows updates, and sleep/wake cycles. Once it stops, Cowork hangs or fails to connect even though everything else looks fine. This is the most common Cowork issue on Windows and the fastest one to fix.",[14,489,490],{},"Open PowerShell as Administrator and run:",[135,492,496],{"className":493,"code":494,"language":495,"meta":143,"style":143},"language-powershell shiki shiki-themes github-light","sc start CoworkVMService\n","powershell",[24,497,498],{"__ignoreMap":143},[499,500,503],"span",{"class":501,"line":502},"line",1,[499,504,494],{},[14,506,507,508,511],{},"If that fixes Cowork until the next reboot, make it stick by switching the service to automatic startup (mind the space after ",[24,509,510],{},"=","):",[135,513,515],{"className":493,"code":514,"language":495,"meta":143,"style":143},"sc config CoworkVMService start= auto\n",[24,516,517],{"__ignoreMap":143},[499,518,519],{"class":501,"line":502},[499,520,514],{},[14,522,523,524,527],{},"If ",[24,525,526],{},"sc start"," returns \"service not found,\" skip to the install-related sections below — your Cowork installation may be broken. Otherwise, restart Claude Desktop and check the Cowork tab.",[39,529,531],{"id":530},"check-your-system-before-you-debug","Check your system before you debug",[14,533,534,535,538,539,542,543,546],{},"Anthropic ships a downloadable ",[53,536,537],{},"Cowork readiness checker"," linked from the \"Get started with Claude Cowork\" article in Anthropic's help center (separate utility, not part of Claude Desktop). Run it first — it reports whether your machine has the Hyper-V components Cowork needs. Caveat: on Windows 11 Home it can falsely report \"ready\" because Windows 11 internally still reports as ",[24,540,541],{},"10.0"," and the checker misidentifies the OS (GitHub #50621). If the checker says ready but Cowork won't load, run ",[24,544,545],{},"Get-Service vmms"," in PowerShell — if that service doesn't exist, you're on Home and Cowork won't work.",[39,548,550],{"id":549},"the-five-ways-cowork-breaks-on-windows","The Five Ways Cowork Breaks on Windows",[14,552,553],{},"The problems aren't random. They fall into five distinct patterns, and knowing which one you're hitting is half the battle.",[351,555,557],{"id":556},"_1-the-missing-tab-yukonsilver-bug","1. The Missing Tab (yukonSilver bug)",[14,559,560],{},"You install Claude Desktop, open it, and the Cowork tab simply isn't there. Only \"Chat\" shows up. This is the \"yukonSilver not supported\" bug, tracked in GitHub issues #25136, #32004, and #32837. Claude's internal platform detection incorrectly marks your system as incompatible, even when all virtualization features are enabled.",[351,562,564],{"id":563},"_2-the-infinite-setup-spinner","2. The Infinite Setup Spinner",[14,566,567],{},"The Cowork tab appears, but clicking it shows \"Setting up Claude's workspace\" with a loading bar stuck at 80 to 90%. It never completes. Users have reported leaving it running for 12+ hours with no progress. No error message. Just spinning.",[351,569,571],{"id":570},"_3-the-api-connection-failure","3. The API Connection Failure",[14,573,574],{},"The workspace starts but can't reach Claude's API. You get \"Cannot connect to Claude API from workspace\" or its Japanese equivalent. This was a day-one launch bug on Windows 11 Home and has resurfaced multiple times since.",[351,576,578],{"id":577},"_4-the-network-conflict","4. The Network Conflict",[14,580,581],{},"Cowork uses a hardcoded network range (172.16.0.0/24) for its internal NAT. If your home network, corporate VPN, or another VM tool uses the same range, Cowork's VM can't reach the internet. Worse, it can break your WSL2 and Docker networking in the process.",[351,583,585],{"id":584},"_5-the-update-regression","5. The Update Regression",[14,587,588],{},"Auto-updates have introduced Cowork-breaking regressions more than once. The most-reported example was v1.1.5749 on March 9, 2026, which broke working installs and required a patch release to recover. Anthropic has since shipped multiple updates; if you're stuck on a known-bad version, updating to the current Claude Desktop release is usually the fix.",[14,590,591],{},[46,592],{"alt":593,"src":594},"The five ways Claude Cowork breaks on Windows: missing tab, infinite spinner, API failure, network conflict, and update regression","/img/blog/claude-cowork-not-working-windows-five-bugs.jpg",[39,596,598],{"id":597},"the-windows-home-problem-that-anthropic-still-hasnt-documented","The Windows Home Problem That Anthropic Still Hasn't Documented",[14,600,601],{},"This is where it gets messy.",[14,603,604],{},"Claude Cowork runs inside a lightweight Hyper-V virtual machine on your Windows machine. That's how it creates its sandboxed environment for file access and code execution. The problem? Windows 11 Home doesn't include the full Hyper-V stack.",[14,606,607],{},"Home edition has Virtual Machine Platform and Windows Hypervisor Platform. But it's missing the vmms (Virtual Machine Management) service that Cowork's VM requires. Without it, the VM either fails silently or throws a cryptic \"Plan9 mount failed: bad address\" error.",[14,609,610],{},"At least seven separate GitHub issues have been filed by Windows Home users who spent hours troubleshooting before discovering that their Windows edition simply can't run Cowork. One user explicitly noted they \"subscribed to Max specifically to use this feature\" and only discovered the incompatibility after paying.",[14,612,613],{},"A documentation request (GitHub issue #27906) was filed in February asking Anthropic to add this information clearly. That issue is now closed and the help center's deployment article has been updated to be more explicit about edition requirements.",[14,615,616,617,619,620,622],{},"The quickest check is to open PowerShell and run ",[24,618,545],{},". If the service isn't found, Cowork won't work without upgrading to Windows 11 Pro or Enterprise. Don't rely on Anthropic's readiness checker alone — GitHub #50621 documents that it falsely reports \"ready\" on Windows 11 Home because Windows 11 internally still reports its version as ",[24,621,541],{}," and the checker misclassifies the OS. A handful of community reports describe partial functionality on Home, but the official position is that Cowork requires the Pro/Enterprise Hyper-V stack.",[39,624,626],{"id":625},"the-yukonsilver-bug-and-why-your-pro-machine-still-fails","The \"yukonSilver\" Bug and Why Your Pro Machine Still Fails",[14,628,629],{},"Stay with me here, because this one is especially frustrating.",[14,631,632],{},"Even if you're running Windows 11 Pro with every virtualization feature enabled (Hyper-V, VMP, WHP, WSL2), you might still see the Cowork tab missing entirely. The logs will show \"yukonSilver not supported (status=unsupported)\" followed by the VM bundle cleanup routine running instead of the actual VM boot.",[14,634,635],{},"\"yukonSilver\" is Claude's internal codename for its VM configuration on Windows. The bug is in the platform detection logic: it incorrectly classifies fully capable x64 Windows 11 Pro systems as unsupported.",[14,637,638,639,642],{},"But that's not even the real problem. The installer also creates a Windows service called CoworkVMService, and this service sometimes becomes impossible to remove. Running ",[24,640,641],{},"sc.exe delete CoworkVMService"," as Administrator returns \"Access denied.\" The service blocks clean reinstalls and creates a circular failure where you can't fix the problem and you can't start fresh.",[14,644,645,646,649],{},"The documented workaround from community debugging: manually run ",[24,647,648],{},"Add-AppxPackage"," as the target user to install the MSIX package correctly for your account. It's a PowerShell command that most of Cowork's target audience (non-developers) would never discover on their own.",[351,651,653],{"id":652},"squirrel-vs-msix-which-installer-do-you-have","Squirrel vs. MSIX: which installer do you have?",[14,655,656,657,660,661,664],{},"Anthropic switched Claude Desktop on Windows from a Squirrel ",[24,658,659],{},".exe"," installer to an MSIX/Microsoft Store package around February 10-13, 2026. If you installed Claude Desktop before that, you have the Squirrel build, and the in-app \"Reinstall\" button can silently fail (tracked in GitHub issues #25162, #25385, #26457; error code ",[24,662,663],{},"0x80073CFA"," in some logs). The fix is a manual uninstall via \"Add or remove programs,\" followed by downloading the fresh MSIX from the official Claude download page. Note that MSIX installs also require Windows \"Sideload apps\" / \"Trusted App Installs\" to be enabled — without it, the MSIX install fails before it starts.",[14,666,667],{},"As one developer debugging the issue put it: \"Cowork is marketed at the people least equipped to debug it when it breaks.\"",[14,669,670,671,675],{},"If you've been running into similar infrastructure headaches with AI agents and want something that works out of the box, our ",[158,672,674],{"href":673},"/compare/self-hosted","comparison of self-hosted vs managed OpenClaw deployments"," covers why some teams are moving away from local setups entirely.",[39,677,679],{"id":678},"the-network-bug-that-breaks-docker-too","The Network Bug That Breaks Docker Too",[14,681,682],{},"Here's what nobody tells you about Cowork's networking on Windows.",[14,684,685],{},"Cowork creates its own Hyper-V virtual switch and NAT network. It's separate from WSL2's networking and separate from Docker Desktop's networking. Three different tenants sharing the same hypervisor, each with their own plumbing.",[14,687,688],{},"The specific failure: Cowork creates an HNS (Host Network Service) network called \"cowork-vm-nat\" but sometimes fails to create the corresponding WinNAT rule. The HNS network exists, but there's no NAT translation. The VM boots, but it has no internet access.",[14,690,691],{},"And in a particularly fun bug, Cowork's virtual network has been reported to permanently break WSL2's internet connectivity until you manually find and delete the offending network configuration using PowerShell HNS diagnostic tools.",[14,693,694],{},"The fix, discovered by community members, involves stopping all Claude processes, killing the Cowork VM via hcsdiag, removing the broken HNS network, and recreating it on a non-conflicting subnet like 172.24.0.0/24 or 10.200.0.0/24.",[14,696,697],{},"This is three PowerShell commands for someone who knows what they're doing. For someone who just wanted to organize their Downloads folder with AI, it's a wall.",[14,699,700],{},[46,701],{"alt":702,"src":703},"Cowork network conflict diagram showing Hyper-V NAT, WSL2, and Docker competing on the same subnet","/img/blog/claude-cowork-not-working-windows-network-conflict.jpg",[39,705,707],{"id":706},"what-actually-fixes-each-bug-quick-reference","What Actually Fixes Each Bug (Quick Reference)",[351,709,711],{"id":710},"missing-cowork-tab-yukonsilver-bug","Missing Cowork Tab (yukonSilver bug)",[14,713,714],{},"First, confirm you're not on Windows Home. If you're on Pro or Enterprise and still don't see the tab, fully uninstall Claude Desktop, remove the leftover service, and clear residual files before reinstalling:",[135,716,718],{"className":493,"code":717,"language":495,"meta":143,"style":143},"sc.exe stop CoworkVMService\nsc.exe delete CoworkVMService\nRemove-Item -Recurse \"$env:APPDATA\\Claude\"\nRemove-Item -Recurse \"$env:LOCALAPPDATA\\Packages\\Claude_*\"\n",[24,719,720,725,730,735],{"__ignoreMap":143},[499,721,722],{"class":501,"line":502},[499,723,724],{},"sc.exe stop CoworkVMService\n",[499,726,727],{"class":501,"line":423},[499,728,729],{},"sc.exe delete CoworkVMService\n",[499,731,732],{"class":501,"line":434},[499,733,734],{},"Remove-Item -Recurse \"$env:APPDATA\\Claude\"\n",[499,736,738],{"class":501,"line":737},4,[499,739,740],{},"Remove-Item -Recurse \"$env:LOCALAPPDATA\\Packages\\Claude_*\"\n",[14,742,743],{},"Then reinstall fresh from the official Claude download page.",[351,745,747],{"id":746},"infinite-setup-spinner","Infinite Setup Spinner",[14,749,750,751,754],{},"Two common causes here. First, the VM download itself. Look in ",[24,752,753],{},"%APPDATA%\\Claude\\vm_bundles\\"," — if the directory is empty or incomplete, your download was interrupted and a clean reinstall usually resolves it.",[14,756,757,758,761,762,765,766,769,770,773,774,777],{},"Second, the ",[53,759,760],{},"cross-drive storage path bug"," (GitHub #36642, #30584, #37754). Cowork writes ",[24,763,764],{},"rootfs.vhdx"," to ",[24,767,768],{},"C:\\Windows\\Temp"," first and then tries to rename it into its final location. If Windows \"Where new content is saved\" sends user data to a non-C: drive, that rename crosses devices and Node.js throws ",[24,771,772],{},"EXDEV: cross-device link not permitted",". Symptom: the spinner hangs forever with no visible error. Fix: open ",[53,775,776],{},"Settings → System → Storage → Advanced storage settings → Where new content is saved",", switch \"New apps\" back to the C: drive, and retry.",[14,779,780],{},"If the spinner persists on Windows Home, it's the Hyper-V incompatibility and there's no fix short of upgrading your edition.",[351,782,784],{"id":783},"api-connection-failure","API Connection Failure",[14,786,787],{},"Disable your VPN temporarily (fully quit, don't just disconnect). Check whether your network uses the 172.16.0.0/24 range. If Chat works but Cowork doesn't, the problem is the VM's network stack, not your internet. Update to the latest Claude Desktop — v1.1.4328 or higher specifically addressed early API connection bugs.",[351,789,791],{"id":790},"network-conflict","Network Conflict",[14,793,794],{},"Check whether Cowork's HNS network exists but the NAT rule doesn't:",[135,796,798],{"className":493,"code":797,"language":495,"meta":143,"style":143},"Get-NetNat\nGet-HnsNetwork | Where-Object {$_.Name -eq \"cowork-vm-nat\"}\n",[24,799,800,805],{"__ignoreMap":143},[499,801,802],{"class":501,"line":502},[499,803,804],{},"Get-NetNat\n",[499,806,807],{"class":501,"line":423},[499,808,809],{},"Get-HnsNetwork | Where-Object {$_.Name -eq \"cowork-vm-nat\"}\n",[14,811,523,812,815],{},[24,813,814],{},"Get-NetNat"," is empty but the HNS query returns a result, you're in the \"missing NAT rule\" state. Remove the broken network and recreate it on a non-conflicting subnet like 172.24.0.0/24 or 10.200.0.0/24.",[351,817,819],{"id":818},"update-regression-v115749","Update Regression (v1.1.5749)",[14,821,822],{},"If Cowork broke after the March 9 update, there's no user-side fix. Update to the latest Claude Desktop release — Anthropic has shipped multiple patches since.",[14,824,825,826,830],{},"If all of this sounds like a lot of infrastructure debugging for a tool that's supposed to \"just work,\" that's because it is. This is the kind of operational friction we built ",[158,827,829],{"href":828},"/","BetterClaw"," to eliminate. Your OpenClaw agent runs on managed infrastructure: no local VMs, no Hyper-V, no NAT conflicts. $19/month, BYOK, first deploy in ~60 seconds.",[39,832,834],{"id":833},"cowork-vs-a-managed-agent-pick-what-matches-your-job","Cowork vs. a managed agent: pick what matches your job",[14,836,837],{},"Cowork is a desktop co-pilot. It runs a local Hyper-V VM, which is why every Windows edition quirk, network conflict, and update regression becomes a potential failure point. If you need a co-pilot you sit beside, that trade-off makes sense.",[14,839,840,841,845,846,195],{},"If you need an always-on agent that handles tasks across messaging platforms while your computer is asleep, the architecture has to be different. ",[158,842,844],{"href":843},"/openclaw-hosting","Managed OpenClaw hosting"," runs your agent on cloud infrastructure with Slack, Discord, WhatsApp, and 15+ channels. No local VM, no Hyper-V, no PowerShell on a Tuesday night. $19/agent/month, BYOK, first deploy in ~60 seconds. ",[158,847,849],{"href":341,"rel":848},[343],"Start free",[39,851,349],{"id":348},[14,853,854],{},[53,855,856],{},"Why is Claude Cowork not working on my Windows machine?",[14,858,859],{},"Top causes: CoworkVMService stopped after reboot or sleep, Windows Home edition missing the full Hyper-V stack, the \"yukonSilver\" platform detection bug, network conflicts with VPNs or Docker on 172.16.0.0/24, or a corrupted install from the old Squirrel package. Check your Windows edition, then your VM service state, then the Claude Code GitHub issues for your exact error.",[14,861,862],{},[53,863,864],{},"How do I restart CoworkVMService on Windows?",[14,866,867,868,871,872,875,876,878],{},"Open PowerShell as Administrator and run ",[24,869,870],{},"sc start CoworkVMService"," to start it for the current session. To make it survive reboots, run ",[24,873,874],{},"sc config CoworkVMService start= auto"," (mind the space after ",[24,877,510],{},"). The service ships with startup type Manual, so it stops after reboots, Windows updates, and sleep/wake cycles. This is the single most common Cowork fix on Windows.",[14,880,881],{},[53,882,883],{},"Does Claude Cowork work on Windows 11 Home?",[14,885,886,887,890,891,893],{},"No, officially. Cowork requires the Hyper-V ",[24,888,889],{},"vmms"," service, which Home editions lack. Anthropic's readiness checker can falsely report Home as \"ready\" (it misidentifies the OS), so don't trust it alone — run ",[24,892,545],{}," in PowerShell. If it's missing, upgrade to Windows 11 Pro or Enterprise.",[14,895,896],{},[53,897,898],{},"How do I fix the \"yukonSilver not supported\" error in Claude Cowork?",[14,900,901,902,905,906,909],{},"This is a platform detection bug on Claude's side, still open as of May 2026. The workaround: fully uninstall Claude Desktop, stop and delete CoworkVMService via elevated PowerShell, clear ",[24,903,904],{},"%APPDATA%\\Claude"," and the ",[24,907,908],{},"%LOCALAPPDATA%\\Packages\\Claude_*"," folder, then reinstall fresh from the official download.",[14,911,912],{},[53,913,914],{},"Is Claude Cowork on Windows stable enough for daily use in 2026?",[14,916,917],{},"Cowork went GA in April 2026, but Windows is still the rougher platform. The yukonSilver bug remains open, the CoworkVMService Manual-startup behavior catches users after every reboot, and update regressions appear periodically. Fine for desktop tasks if your system is compatible. For workloads where downtime means lost work, a managed agent is more reliable.",[919,920,921],"style",{},"html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}",{"title":143,"searchDepth":423,"depth":423,"links":923},[924,925,926,933,934,937,938,945,946],{"id":479,"depth":423,"text":480},{"id":530,"depth":423,"text":531},{"id":549,"depth":423,"text":550,"children":927},[928,929,930,931,932],{"id":556,"depth":434,"text":557},{"id":563,"depth":434,"text":564},{"id":570,"depth":434,"text":571},{"id":577,"depth":434,"text":578},{"id":584,"depth":434,"text":585},{"id":597,"depth":423,"text":598},{"id":625,"depth":423,"text":626,"children":935},[936],{"id":652,"depth":434,"text":653},{"id":678,"depth":423,"text":679},{"id":706,"depth":423,"text":707,"children":939},[940,941,942,943,944],{"id":710,"depth":434,"text":711},{"id":746,"depth":434,"text":747},{"id":783,"depth":434,"text":784},{"id":790,"depth":434,"text":791},{"id":818,"depth":434,"text":819},{"id":833,"depth":423,"text":834},{"id":348,"depth":423,"text":349},"2026-03-27","Claude Cowork broken on Windows? Covers 5 failure modes: yukonSilver bug, Windows Home limits, VM service stops, network conflicts, and update regressions. Step-by-step fixes.","/img/blog/claude-cowork-not-working-windows.jpg",{},"/blog/claude-cowork-not-working-windows","14 min read",{"title":465,"description":948},"Claude Cowork Not Working on Windows? 5 Fixes (2026)","blog/claude-cowork-not-working-windows",[957,958,959,960,961,962],"Claude Cowork not working Windows","Cowork Windows bugs","yukonSilver error","Claude Cowork Windows fix","Cowork Hyper-V","Cowork Windows Home","Sq06kygerdAX6Y2Mxv3OPy2yjFOapLcXpUWGeKLwStI",{"id":965,"title":966,"author":967,"body":968,"category":439,"date":1346,"description":1347,"extension":442,"featured":443,"image":1348,"imageHeight":445,"imageWidth":445,"meta":1349,"navigation":447,"path":1350,"readingTime":1351,"seo":1352,"seoTitle":1353,"stem":1354,"tags":1355,"updatedDate":1346,"__hash__":1362},"blog/blog/openclaw-agent-hallucinating-fix.md","OpenClaw Agent Hallucinating? Why It's Describing Tasks Instead of Doing Them",{"name":7,"role":8,"avatar":9},{"type":11,"value":969,"toc":1334},[970,976,979,982,985,988,992,995,998,1001,1005,1008,1011,1022,1032,1038,1042,1045,1048,1058,1064,1068,1071,1074,1084,1090,1094,1097,1100,1118,1124,1128,1135,1144,1152,1158,1164,1168,1171,1177,1183,1189,1200,1209,1216,1220,1223,1226,1229,1235,1243,1245,1250,1256,1261,1264,1269,1272,1277,1280,1285,1294,1298],[14,971,972],{},[973,974,975],"em",{},"Your agent says \"I've searched the web for you\" but didn't actually search. Here's the specific reason and the fix for each cause.",[14,977,978],{},"I asked my OpenClaw agent to check the weather in London. It responded with a detailed forecast: 14 degrees, partly cloudy, 60% chance of rain in the afternoon.",[14,980,981],{},"The forecast was completely wrong. Not because the weather API was broken. Because the agent never called the weather API. It generated a plausible-sounding forecast from its training data and presented it as if it had just looked it up.",[14,983,984],{},"This is the most frustrating OpenClaw behavior: the agent describes doing something without actually doing it. It says \"I've searched for that\" without searching. It says \"I've checked your calendar\" without checking. It writes a confident response that looks like it came from a tool call but was entirely fabricated.",[14,986,987],{},"Here's what nobody tells you: this isn't a bug in OpenClaw. It's a predictable failure mode with five specific causes, each with a different fix.",[39,989,991],{"id":990},"the-difference-between-hallucinating-and-executing","The difference between hallucinating and executing",[14,993,994],{},"When your OpenClaw agent properly executes a task, the process looks like this: you send a message, the model decides which tool to call, OpenClaw executes the tool, the tool returns real data, and the model generates a response based on that real data.",[14,996,997],{},"When the agent hallucinates a task, the process looks like this: you send a message, the model skips the tool call entirely, and generates a response that looks like it used a tool but didn't. No tool was called. No real data was retrieved. The response is pure fabrication dressed up as fact.",[14,999,1000],{},"The scary part is that both responses look identical to you. The agent doesn't say \"I'm guessing.\" It presents the hallucinated answer with the same confidence as a real one.",[39,1002,1004],{"id":1003},"cause-1-your-model-doesnt-support-tool-calling","Cause 1: Your model doesn't support tool calling",[14,1006,1007],{},"This is the most common cause and the easiest to fix.",[14,1009,1010],{},"Not every AI model can call tools. Tool calling is a specific capability that models must be trained for. If your model doesn't support it, the agent has no way to execute tools. It does the next best thing: it describes what it would do if it could.",[14,1012,1013,1014,1017,1018,1021],{},"This especially affects Ollama users running local models. Models like ",[24,1015,1016],{},"phi3:mini",", ",[24,1019,1020],{},"qwen2.5:3b",", and other small models lack tool calling support entirely. Even models that support tool calling through Ollama have issues because of a streaming bug (GitHub Issue #5769) that drops tool call responses.",[14,1023,1024,1026,1027,1031],{},[53,1025,91],{}," Switch to a model that supports tool calling. For cloud providers: Claude Sonnet, GPT-4o, DeepSeek, and Gemini all support tool calling reliably. For the ",[158,1028,1030],{"href":1029},"/blog/openclaw-model-does-not-support-tools","full breakdown of which models support tools and which don't",", our model compatibility guide covers every common model.",[14,1033,1034],{},[46,1035],{"alt":1036,"src":1037},"OpenClaw model tool calling support matrix showing which cloud and local models work with tools","/img/blog/openclaw-agent-hallucinating-fix-models.jpg",[39,1039,1041],{"id":1040},"cause-2-docker-isnt-running-so-sandboxed-execution-fails-silently","Cause 2: Docker isn't running (so sandboxed execution fails silently)",[14,1043,1044],{},"OpenClaw uses Docker containers for sandboxed code execution and some tool operations. If Docker Desktop isn't running (on Mac/Windows) or the Docker daemon isn't active (on Linux/VPS), tool calls that require sandboxed execution fail silently.",[14,1046,1047],{},"Here's the weird part. The agent doesn't always tell you Docker failed. Instead, it falls back to generating a response without the tool, making it look like it executed the task when it couldn't.",[14,1049,1050,1052,1053,1057],{},[53,1051,91],{}," Make sure Docker is running before starting OpenClaw. On Mac/Windows, check for the Docker Desktop whale icon in the system tray. On Linux, verify the Docker daemon is active. For the ",[158,1054,1056],{"href":1055},"/blog/openclaw-docker-troubleshooting","complete Docker troubleshooting guide",", our guide covers the eight most common Docker errors and their fixes.",[14,1059,1060],{},[46,1061],{"alt":1062,"src":1063},"OpenClaw Docker dependency diagram showing how sandboxed tools fail silently when Docker daemon is down","/img/blog/openclaw-agent-hallucinating-fix-docker.jpg",[39,1065,1067],{"id":1066},"cause-3-the-skill-you-think-is-installed-isnt-actually-active","Cause 3: The skill you think is installed isn't actually active",[14,1069,1070],{},"You installed a web search skill last week. You ask the agent to search something. It generates a fake search result instead of actually searching.",[14,1072,1073],{},"The skill might have been deactivated by a recent OpenClaw update. It might have failed validation after a version change. It might be installed globally but not in the current workspace. OpenClaw doesn't always tell you when a skill goes inactive.",[14,1075,1076,1078,1079,1083],{},[53,1077,91],{}," Check your installed skills. Ask the agent to list its available tools. If the skill you expect isn't in the list, reinstall it. After any OpenClaw update, verify your skills are still active. For the ",[158,1080,1082],{"href":1081},"/blog/clawhub-skills-directory","skill audit process including how to verify what's installed",", our skills guide covers the verification steps.",[14,1085,1086],{},[46,1087],{"alt":1088,"src":1089},"OpenClaw skill verification flow showing how to check active skills, reinstall after updates, and confirm tool availability","/img/blog/openclaw-agent-hallucinating-fix-skills.jpg",[39,1091,1093],{"id":1092},"cause-4-the-agent-is-stuck-in-a-reasoning-loop","Cause 4: The agent is stuck in a reasoning loop",[14,1095,1096],{},"Sometimes the agent enters a loop where it tries to call a tool, encounters an error, retries, encounters the same error, and eventually gives up and generates a response without the tool. From your perspective, you asked a question and got an answer. You didn't see the five failed tool attempts that happened behind the scenes.",[14,1098,1099],{},"The agent doesn't announce that it gave up. It just... answers. With fabricated data. As if nothing went wrong.",[14,1101,1102,1104,1105,1108,1109,1112,1113,1117],{},[53,1103,91],{}," Check the gateway logs for repeated tool call errors. If you see the same tool being called and failing multiple times, there's a skill error or a configuration problem causing the loop. Set ",[24,1106,1107],{},"maxIterations"," to 10-15 in your config to prevent infinite retries. Use ",[24,1110,1111],{},"/new"," to clear the session state. For the ",[158,1114,1116],{"href":1115},"/blog/openclaw-agent-stuck-in-loop","complete guide to diagnosing agent loops",", our loop troubleshooting post covers the specific patterns.",[14,1119,1120],{},[46,1121],{"alt":1122,"src":1123},"OpenClaw silent retry loop showing how repeated tool failures lead to fabricated responses without user-visible errors","/img/blog/openclaw-agent-hallucinating-fix-loop.jpg",[39,1125,1127],{"id":1126},"cause-5-your-soulmd-is-conflicting-with-tool-use","Cause 5: Your SOUL.md is conflicting with tool use",[14,1129,1130,1131,1134],{},"This is the subtlest cause. If your ",[24,1132,1133],{},"SOUL.md"," contains instructions that discourage or limit tool use (\"answer from your knowledge first,\" \"don't use tools unless necessary,\" \"respond quickly without external lookups\"), the model may interpret these as reasons to skip tool calls and generate responses from its training data instead.",[14,1136,1137,1138,1140,1141,1143],{},"The model follows your ",[24,1139,1133],{},". If the ",[24,1142,1133],{}," suggests that responding quickly from knowledge is preferred over using tools, the model will do exactly that. Even when using tools would give a better answer.",[14,1145,1146,1148,1149,1151],{},[53,1147,91],{}," Review your ",[24,1150,1133],{}," for any instructions that could be interpreted as \"don't use tools.\" Remove or clarify them. If you want the agent to always use tools for certain types of queries (web search for current information, calendar checks for scheduling), add explicit instructions: \"Always use web search for questions about current events, prices, or availability. Never guess when a tool can provide the real answer.\"",[14,1153,1154,1155,1157],{},"When your agent hallucinates tool use, it's not broken. It's choosing not to use tools because of one of five specific reasons: the model can't call tools, Docker isn't running, the skill is inactive, the tool is failing silently, or your ",[24,1156,1133],{}," discouraged tool use. Fix the specific cause. The hallucination stops.",[14,1159,1160],{},[46,1161],{"alt":1162,"src":1163},"OpenClaw SOUL.md tool use conflicts showing instructions that accidentally discourage tool calling and how to rewrite them","/img/blog/openclaw-agent-hallucinating-fix-soulmd.jpg",[39,1165,1167],{"id":1166},"the-quick-diagnostic-run-this-in-2-minutes","The quick diagnostic (run this in 2 minutes)",[14,1169,1170],{},"When your agent describes a task instead of doing it, check these five things in this order.",[14,1172,1173,1176],{},[53,1174,1175],{},"First",", verify your model supports tool calling. If you're on Ollama, this is probably the issue. Switch to a cloud provider temporarily to test.",[14,1178,1179,1182],{},[53,1180,1181],{},"Second",", verify Docker is running. Check the system tray (Mac/Windows) or daemon status (Linux).",[14,1184,1185,1188],{},[53,1186,1187],{},"Third",", ask the agent to list its available tools. If the tool you expected isn't listed, reinstall the skill.",[14,1190,1191,1194,1195,1197,1198,195],{},[53,1192,1193],{},"Fourth",", check the gateway logs for repeated tool call errors. If you see retries, set ",[24,1196,1107],{}," and use ",[24,1199,1111],{},[14,1201,1202,1205,1206,1208],{},[53,1203,1204],{},"Fifth",", review your ",[24,1207,1133],{}," for any instructions that discourage tool use.",[14,1210,1211,1212,1215],{},"If debugging tool calling failures and Docker dependencies isn't how you want to spend your afternoon, ",[158,1213,1214],{"href":843},"Better Claw handles tool execution"," with Docker-sandboxed execution built into the platform. $19/month per agent, BYOK with 28+ providers. Every model we support has working tool calling. Skills execute in sandboxed containers. No silent failures. No hallucinated tool use.",[39,1217,1219],{"id":1218},"why-this-matters-more-than-most-people-realize","Why this matters more than most people realize",[14,1221,1222],{},"Here's the uncomfortable truth about agent hallucination.",[14,1224,1225],{},"When your agent hallucinates a web search and gives you wrong information, you can probably tell. When it hallucinates a calendar check and tells you your afternoon is free (when it isn't), the consequences are more serious. When it hallucinates a file operation and tells you it saved something (when it didn't), you lose data.",[14,1227,1228],{},"The Meta researcher Summer Yue incident (agent mass-deleting emails while ignoring stop commands) is the extreme case. But the everyday case is agents that claim to have done things they didn't do. Not maliciously. Just because the tool call failed and the model covered the gap with a confident-sounding response.",[14,1230,1231,1232,1234],{},"The fix isn't to distrust your agent. The fix is to ensure tool calling actually works (right model, Docker running, skills active, no loops, clear ",[24,1233,1133],{},") and to verify important actions by checking the results independently until you trust the pipeline.",[14,1236,1237,1238,1242],{},"If you want an agent where tool calls execute reliably and failures surface clearly instead of being masked by hallucination, ",[158,1239,1241],{"href":341,"rel":1240},[343],"give Better Claw a try",". $19/month per agent, BYOK with 28+ providers. 60-second deploy. Health monitoring catches tool execution failures before they become hallucinated answers.",[39,1244,349],{"id":348},[14,1246,1247],{},[53,1248,1249],{},"Why does my OpenClaw agent describe tasks instead of executing them?",[14,1251,1252,1253,1255],{},"The most common cause is that your model doesn't support tool calling (especially local Ollama models under 7B parameters). Other causes: Docker not running (sandboxed execution fails silently), the required skill being inactive after an update, the agent stuck in a retry loop, or ",[24,1254,1133],{}," instructions that discourage tool use. The agent generates a confident response from its training data instead of admitting the tool call failed.",[14,1257,1258],{},[53,1259,1260],{},"How do I know if my OpenClaw agent is hallucinating?",[14,1262,1263],{},"Check the gateway logs for tool call entries. If your agent claims to have searched the web but the logs show no web search tool call, the response was hallucinated. You can also test by asking for verifiable information (today's date, current weather, a specific fact you can check). If the answer is wrong or outdated, the agent likely generated it from training data rather than using a tool.",[14,1265,1266],{},[53,1267,1268],{},"Which models support tool calling in OpenClaw?",[14,1270,1271],{},"Cloud models with reliable tool calling: Claude Sonnet, Claude Opus, GPT-4o, DeepSeek, Gemini Pro. Local Ollama models with tool calling support (but affected by streaming bug): hermes-2-pro, mistral:7b, qwen3:8b+, llama3.1:8b+. Models without tool calling: phi3:mini, qwen2.5:3b, and most small quantized models. Cloud providers have the most reliable tool execution because their streaming implementation correctly returns tool call responses.",[14,1273,1274],{},[53,1275,1276],{},"Does Docker need to be running for OpenClaw tools to work?",[14,1278,1279],{},"For skills that use sandboxed execution (code execution, browser automation, some file operations), yes. Docker provides the container environment where these tools run safely. If Docker isn't running, these tool calls fail silently and the agent may hallucinate a response instead. Always verify Docker is running before starting your OpenClaw gateway. Not all tools require Docker (simple API calls, web search through external services), but many core capabilities do.",[14,1281,1282],{},[53,1283,1284],{},"How do I stop my OpenClaw agent from making up information?",[14,1286,1287,1288,1290,1291,1293],{},"Five fixes in order: ensure your model supports tool calling (switch from Ollama to a cloud provider if needed), verify Docker is running, check that required skills are installed and active, set ",[24,1289,1107],{}," to 10-15 to prevent silent retry failures, and review your ",[24,1292,1133],{}," for instructions that might discourage tool use. Add explicit instructions like \"Always use web search for current information. Never guess when a tool can provide the answer.\"",[39,1295,1297],{"id":1296},"related-reading","Related Reading",[1299,1300,1301,1308,1314,1321,1327],"ul",{},[1302,1303,1304,1307],"li",{},[158,1305,1306],{"href":1029},"\"Model Does Not Support Tools\" Fix"," — Tool calling failures by model and provider",[1302,1309,1310,1313],{},[158,1311,1312],{"href":1055},"OpenClaw Docker Troubleshooting Guide"," — Docker errors that cause silent tool failures",[1302,1315,1316,1320],{},[158,1317,1319],{"href":1318},"/blog/openclaw-skill-audit","OpenClaw Skill Audit"," — How to verify which skills are actually active",[1302,1322,1323,1326],{},[158,1324,1325],{"href":1115},"OpenClaw Agent Stuck in Loop"," — Diagnose and fix the silent retry loops",[1302,1328,1329,1333],{},[158,1330,1332],{"href":1331},"/blog/openclaw-soulmd-guide","The OpenClaw SOUL.md Guide"," — Write a system prompt that doesn't discourage tool use",{"title":143,"searchDepth":423,"depth":423,"links":1335},[1336,1337,1338,1339,1340,1341,1342,1343,1344,1345],{"id":990,"depth":423,"text":991},{"id":1003,"depth":423,"text":1004},{"id":1040,"depth":423,"text":1041},{"id":1066,"depth":423,"text":1067},{"id":1092,"depth":423,"text":1093},{"id":1126,"depth":423,"text":1127},{"id":1166,"depth":423,"text":1167},{"id":1218,"depth":423,"text":1219},{"id":348,"depth":423,"text":349},{"id":1296,"depth":423,"text":1297},"2026-04-11","Your OpenClaw agent says it searched the web but didn't. Five causes: wrong model, Docker down, skill inactive, loop, or SOUL.md conflict. Fixes here.","/img/blog/openclaw-agent-hallucinating-fix.jpg",{},"/blog/openclaw-agent-hallucinating-fix","10 min read",{"title":966,"description":1347},"OpenClaw Agent Hallucinating? Not Executing Tasks?","blog/openclaw-agent-hallucinating-fix",[1356,1357,1358,1359,1360,1361],"OpenClaw hallucinating","OpenClaw not executing tasks","OpenClaw tool calling not working","OpenClaw agent making things up","OpenClaw fake responses","OpenClaw agent fix","5vx9y6T-F-bQ0xrEiaxhDb3gEC6jXhgM5n_4AwD7E54",{"id":1364,"title":1365,"author":1366,"body":1367,"category":439,"date":1795,"description":1796,"extension":442,"featured":443,"image":1797,"imageHeight":445,"imageWidth":445,"meta":1798,"navigation":447,"path":1115,"readingTime":1799,"seo":1800,"seoTitle":1801,"stem":1802,"tags":1803,"updatedDate":1810,"__hash__":1811},"blog/blog/openclaw-agent-stuck-in-loop.md","OpenClaw Agent Stuck in Loop? Here's Why You're Burning $25+ in Minutes (And How to Stop It)",{"name":7,"role":8,"avatar":9},{"type":11,"value":1368,"toc":1778},[1369,1382,1387,1390,1393,1396,1399,1402,1405,1409,1412,1415,1418,1421,1424,1427,1430,1436,1440,1443,1446,1449,1452,1455,1458,1461,1464,1468,1475,1479,1482,1485,1489,1492,1500,1504,1507,1513,1517,1520,1523,1526,1534,1537,1540,1544,1547,1557,1563,1569,1579,1586,1590,1597,1600,1606,1609,1613,1616,1619,1622,1625,1628,1635,1639,1642,1653,1659,1665,1672,1676,1679,1682,1685,1688,1691,1699,1701,1706,1709,1714,1717,1722,1731,1736,1739,1744,1747,1749],[14,1370,1371],{},[53,1372,1373,1374,1377,1378,1381],{},"To stop an OpenClaw agent loop, SSH into your server and run ",[24,1375,1376],{},"docker restart openclaw",". Then prevent future loops by setting ",[24,1379,1380],{},"maxIterations: 15"," in your agent config, adding a per-task cost ceiling, and configuring cooldown periods between retries. Agent loops happen when a failed action triggers infinite retry cycles — each burning API tokens.",[14,1383,1384],{},[53,1385,1386],{},"Your agent isn't broken. It's just expensive. Here's what's actually happening when OpenClaw loops, and the fastest way to stop the bleeding.",[14,1388,1389],{},"It was 11:47 PM on a Tuesday. I'd set up an OpenClaw agent to summarize support tickets and push updates to Slack. Simple workflow. Twenty minutes, tops.",[14,1391,1392],{},"I went to bed.",[14,1394,1395],{},"I woke up to a $38 API bill from Anthropic. For one night.",[14,1397,1398],{},"The agent had gotten stuck in a retry loop. Every failed Slack post triggered another reasoning cycle. Every reasoning cycle packed more context into the prompt. Every prompt burned more tokens. For six hours straight, my agent was essentially arguing with itself about why a Slack webhook URL was wrong, spending real money on every single turn of that argument.",[14,1400,1401],{},"If you're running OpenClaw and you've seen your API costs spike without explanation, you're not alone. And this isn't a bug. It's a design reality of how autonomous agents work.",[14,1403,1404],{},"Here's what's actually going on.",[39,1406,1408],{"id":1407},"why-your-openclaw-agent-gets-stuck-its-not-what-you-think","Why Your OpenClaw Agent Gets Stuck (It's Not What You Think)",[14,1410,1411],{},"Most people assume a looping agent means something is misconfigured. Bad YAML. Wrong API key. Broken skill file.",[14,1413,1414],{},"Sometimes, yes. But the more common cause is subtler and more expensive.",[14,1416,1417],{},"OpenClaw agents operate on a reason-act-observe loop. The agent reads its context, decides what to do, takes an action, observes the result, and then reasons again. This is the core pattern behind every agent framework, not just OpenClaw.",[14,1419,1420],{},"The problem starts when the \"observe\" step returns ambiguous feedback.",[14,1422,1423],{},"Think about it. If a tool call returns \"request failed, please try again,\" the agent should try again. That's what it's designed to do. It's being a good agent. But without explicit limits on how many times it retries, or any awareness of how much each retry costs, it will keep trying forever.",[14,1425,1426],{},"Research from AWS shows that agents can loop hundreds of times without delivering a single useful result when tool feedback is vague. The agent keeps calling the same tool with slightly different parameters, convinced the next attempt will work.",[14,1428,1429],{},"And every single one of those attempts costs tokens.",[14,1431,1432],{},[46,1433],{"alt":1434,"src":1435},"OpenClaw reason-act-observe loop diagram showing how ambiguous tool feedback triggers infinite retries","/img/blog/openclaw-agent-stuck-in-loop-reason-loop.jpg",[39,1437,1439],{"id":1438},"the-math-that-should-scare-you","The Math That Should Scare You",[14,1441,1442],{},"Let's do some quick napkin math on what an OpenClaw loop actually costs.",[14,1444,1445],{},"Say your agent is running Claude Sonnet. Each reasoning cycle sends the full conversation history plus tool definitions plus the latest observation. That's easily 50,000 to 80,000 input tokens per turn once context starts growing.",[14,1447,1448],{},"At Anthropic's current pricing, that's roughly $0.15 to $0.24 per turn for input tokens alone. Add output tokens and you're looking at $0.20 to $0.35 per reasoning cycle.",[14,1450,1451],{},"Now imagine 100 cycles in an hour. That's $20 to $35 burned on a single stuck task.",[14,1453,1454],{},"Switch to a more powerful model like Claude Opus? The numbers get worse fast. And if your agent is running overnight or over a weekend with no circuit breaker, the math becomes genuinely painful.",[14,1456,1457],{},"A single runaway agent loop can consume your monthly API budget in hours. This isn't hypothetical. It happens to people building with autonomous agents every single week.",[14,1459,1460],{},"One developer recently filed a bug report showing a subagent that burned $350 in 3.5 hours after entering an infinite tool-call loop with 809 consecutive turns. The agent kept reading and re-reading the same files, never concluding its task. Worse, the cost dashboard showed only half the real bill due to a pricing tier mismatch.",[14,1462,1463],{},"This is the risk nobody talks about in the \"just deploy an agent\" tutorials.",[39,1465,1467],{"id":1466},"the-three-loop-patterns-that-drain-your-wallet","The Three Loop Patterns That Drain Your Wallet",[14,1469,1470,1471,1474],{},"Not all loops are created equal. In our experience running managed OpenClaw deployments at ",[158,1472,1473],{"href":828},"Better Claw",", we see three patterns over and over again.",[351,1476,1478],{"id":1477},"_1-the-retry-storm","1. The Retry Storm",[14,1480,1481],{},"A tool call fails. The agent retries. Same error. Retries again. Each retry adds the error message to context, making the prompt longer and more expensive. The agent isn't learning from the failure. It's just paying more to fail again.",[14,1483,1484],{},"This is the most common pattern. It usually comes from external API timeouts, rate limits, or webhook misconfigurations.",[351,1486,1488],{"id":1487},"_2-the-context-avalanche","2. The Context Avalanche",[14,1490,1491],{},"This one is sneakier. The agent successfully calls tools, but each tool returns a massive payload. Full file contents. Entire database query results. Complete API responses. The context window balloons with every turn. Eventually, the agent is spending most of its tokens just reading its own history rather than doing useful work.",[14,1493,1494,1495,1499],{},"If you've looked at ",[158,1496,1498],{"href":1497},"/blog/openclaw-api-costs","how OpenClaw handles API costs",", you know that context management is half the battle.",[351,1501,1503],{"id":1502},"_3-the-verification-loop","3. The Verification Loop",[14,1505,1506],{},"The agent completes a task successfully but then enters an infinite verification cycle. It checks its own work, decides something might be slightly off, \"fixes\" it, checks again, fixes again. Round and round, perfecting something that was already done, burning tokens on what is essentially AI anxiety.",[14,1508,1509],{},[46,1510],{"alt":1511,"src":1512},"Three loop patterns compared: retry storm, context avalanche, and verification loop with cost impact","/img/blog/openclaw-agent-stuck-in-loop-patterns.jpg",[39,1514,1516],{"id":1515},"what-openclaw-doesnt-do-that-you-need-to-do-yourself","What OpenClaw Doesn't Do (That You Need to Do Yourself)",[14,1518,1519],{},"Here's what nobody tells you about self-hosting OpenClaw.",[14,1521,1522],{},"OpenClaw is a powerful agent framework. It handles task execution, skill loading, multi-channel communication, and tool calling really well. But it was designed as a framework, not a managed service. That means certain operational safeguards are left to you.",[14,1524,1525],{},"There's no built-in per-task cost cap. No automatic circuit breaker that kills a loop after N iterations. No alert that fires when token consumption spikes. No rate limiting on the agent's own behavior.",[14,1527,1528,1529,1533],{},"If you're ",[158,1530,1532],{"href":1531},"/blog/openclaw-vps-setup","self-hosting OpenClaw on a VPS",", all of this is your responsibility. You need to configure max retries, set cooldown periods, implement session budgets, and monitor token usage in real time.",[14,1535,1536],{},"The fix itself isn't complicated. A basic circuit breaker config looks something like this: set a max of 3 retries per task, add a 60-second cooldown between failures, cap total actions per session at 50, and kill the agent if it exceeds a dollar threshold per run.",[14,1538,1539],{},"Four rules. That's it. But most people don't add them until after the first surprise bill.",[39,1541,1543],{"id":1542},"how-to-stop-the-bleeding-right-now","How to Stop the Bleeding Right Now",[14,1545,1546],{},"If your agent is stuck in a loop right now, here's what to do.",[14,1548,1549,1552,1553,1556],{},[53,1550,1551],{},"First, kill the process."," Don't wait for it to finish gracefully. Every second it runs is money spent. If you're running in Docker, ",[24,1554,1555],{},"docker stop"," will do it. If you're on a VPS, kill the node process.",[14,1558,1559,1562],{},[53,1560,1561],{},"Second, check your API provider's dashboard."," Look at the token usage for the last few hours. Identify which model was being used and how many requests were made. This tells you the actual damage.",[14,1564,1565,1568],{},[53,1566,1567],{},"Third, look at the agent's conversation history."," Find the point where it started looping. What tool call failed? What was the response? This is your debugging starting point.",[14,1570,1571,1574,1575,1578],{},[53,1572,1573],{},"Fourth, add guardrails before restarting."," Minimum viable guardrails for any OpenClaw deployment: set ",[24,1576,1577],{},"max_retries"," in your agent config, implement a session timeout, and add a cost ceiling per task.",[14,1580,1581,1582,1585],{},"If you want to go deeper on preventing these issues before they start, our guide on ",[158,1583,1584],{"href":160},"OpenClaw best practices"," covers the full configuration approach.",[39,1587,1589],{"id":1588},"the-case-for-not-managing-this-yourself","The Case for Not Managing This Yourself",[14,1591,1592,1593,1596],{},"I'll be direct here. We built ",[158,1594,1473],{"href":1595},"/pricing"," because we got tired of being the human circuit breaker for our own agents.",[14,1598,1599],{},"Every OpenClaw deployment we managed for ourselves had the same lifecycle: set up the agent, it works great for a week, something goes sideways at 2 AM, wake up to a cost spike, spend half a day debugging, add another guardrail, repeat. The agent itself was doing its job. The infrastructure around it was the problem.",[14,1601,1602,1605],{},[158,1603,829],{"href":341,"rel":1604},[343]," runs your OpenClaw agent on managed infrastructure with built-in cost controls, automatic monitoring, and loop detection baked in. $19/month per agent, you bring your own API keys. Your first deploy takes about 60 seconds. We handle the Docker, the uptime, the security patches, and the \"why is my agent spending $50 at 3 AM\" problem.",[14,1607,1608],{},"You handle the interesting part: building the actual workflows your agent runs.",[39,1610,1612],{"id":1611},"the-bigger-picture-why-this-problem-is-getting-worse","The Bigger Picture: Why This Problem Is Getting Worse",[14,1614,1615],{},"Here's something worth thinking about.",[14,1617,1618],{},"As models get smarter, agent loops get more expensive, not less. Newer models have larger context windows, which means a looping agent can accumulate more context before hitting limits. They're also better at generating plausible-sounding reasoning, which means they can loop longer before producing output that looks obviously wrong.",[14,1620,1621],{},"A GPT-4 era agent might loop 50 times before filling its context window. A newer model might loop 500 times in the same window, each turn more expensive than the last.",[14,1623,1624],{},"The industry is moving toward longer-running, more autonomous agents. That's exciting. But it also means the cost of a stuck agent is going up, not down.",[14,1626,1627],{},"The tools for building agents are getting better every month. The tools for operating agents safely are still catching up. That gap is where your API budget disappears.",[14,1629,1630,1631,1634],{},"This is why operational infrastructure matters as much as the agent framework itself. The ",[158,1632,1633],{"href":673},"difference between self-hosted and managed OpenClaw"," isn't just about convenience. It's about whether you have production-grade safeguards running by default or whether you're building them from scratch every time.",[39,1636,1638],{"id":1637},"what-id-tell-someone-just-getting-started","What I'd Tell Someone Just Getting Started",[14,1640,1641],{},"If you're setting up your first OpenClaw agent today, here's what I wish someone had told me.",[14,1643,1644,1647,1648,1652],{},[53,1645,1646],{},"Start with a cheap model for testing."," Use Claude Haiku or GPT-4o-mini while you're iterating on your skill files and task configurations. Switch to a more capable model only after you've confirmed the workflow runs without loops. Our ",[158,1649,1651],{"href":1650},"/blog/openclaw-model-comparison","model comparison guide"," breaks down when each model makes sense.",[14,1654,1655,1658],{},[53,1656,1657],{},"Set cost alerts on your API provider dashboard from day one."," Anthropic, OpenAI, and Google all let you set usage alerts. A $5 daily alert is a simple early warning system.",[14,1660,1661,1664],{},[53,1662,1663],{},"Never leave an agent running overnight without a session timeout."," Just don't. The 30 minutes it takes to add a timeout config will save you hundreds of dollars over the life of your deployment.",[14,1666,1667,1668,1671],{},"And if you'd rather skip the infrastructure headaches entirely and just focus on what your agent does, ",[158,1669,344],{"href":341,"rel":1670},[343],". It's $19/month per agent, BYOK, and your first deploy takes about 60 seconds. We handle the infrastructure. You handle the interesting part.",[39,1673,1675],{"id":1674},"the-real-cost-isnt-the-bill","The Real Cost Isn't the Bill",[14,1677,1678],{},"The thing that actually bothers me about runaway agent loops isn't the money. Money can be recovered.",[14,1680,1681],{},"It's the trust erosion.",[14,1683,1684],{},"Every time an agent loops and burns your budget, it chips away at your confidence in the whole approach. You start second-guessing whether autonomous agents are ready. You add more manual oversight. You reduce the agent's autonomy. And slowly, the thing that was supposed to save you time becomes another system you babysit.",[14,1686,1687],{},"The fix isn't to distrust agents. The fix is to give them proper guardrails so they can be trusted. A well-configured agent with cost caps, retry limits, and monitoring is more autonomous than one you have to watch like a hawk because it might bankrupt you at 3 AM.",[14,1689,1690],{},"Build the guardrails. Trust the agent. Ship the workflow.",[14,1692,1693,1694,1698],{},"Or ",[158,1695,1697],{"href":341,"rel":1696},[343],"let us handle the guardrails"," and skip straight to the good part.",[39,1700,349],{"id":348},[14,1702,1703],{},[53,1704,1705],{},"Why does my OpenClaw agent get stuck in a loop?",[14,1707,1708],{},"OpenClaw agents loop when tool calls return ambiguous or failed responses without clear stop conditions. The agent's reason-act-observe cycle keeps retrying because it's designed to be persistent. Without explicit max-retry limits or circuit breakers configured in your setup, the agent will keep attempting the task indefinitely, burning API tokens on every iteration.",[14,1710,1711],{},[53,1712,1713],{},"How much does an OpenClaw agent loop cost in API fees?",[14,1715,1716],{},"A single stuck loop can cost anywhere from $5 to $50+ per hour depending on your model choice and context size. With Claude Sonnet, expect roughly $0.20 to $0.35 per reasoning cycle. At 100 cycles per hour, that's $20 to $35. One documented case showed a subagent burning $350 in just 3.5 hours during an uncontrolled loop with over 800 consecutive turns.",[14,1718,1719],{},[53,1720,1721],{},"How do I stop an OpenClaw agent that's stuck in a loop right now?",[14,1723,1724,1725,1727,1728,1730],{},"Kill the process immediately. Use ",[24,1726,1555],{}," if running in Docker, or terminate the node process on your VPS. Then check your API provider's usage dashboard to assess the damage. Before restarting, add guardrails: set ",[24,1729,1577],{}," to 3, add a 60-second cooldown between failures, and cap total actions per session at 50.",[14,1732,1733],{},[53,1734,1735],{},"Is BetterClaw worth it compared to self-hosting OpenClaw?",[14,1737,1738],{},"If you value your time and want to avoid surprise API bills, yes. BetterClaw costs $19/month per agent with BYOK (bring your own API keys). You get built-in monitoring, loop detection, and managed infrastructure. Self-hosting is free but requires you to handle Docker maintenance, security patches, uptime monitoring, and building your own cost safeguards from scratch.",[14,1740,1741],{},[53,1742,1743],{},"Can I prevent OpenClaw agent loops without switching to a managed platform?",[14,1745,1746],{},"Absolutely. Set max-retry limits in your agent configuration, implement session timeouts, add per-task cost ceilings, configure cooldown periods between retries, and set up API usage alerts with your provider. These five steps will prevent most runaway loops. The trade-off is that you're responsible for maintaining and updating these safeguards yourself as OpenClaw evolves.",[39,1748,1297],{"id":1296},[1299,1750,1751,1758,1765,1772],{},[1302,1752,1753,1757],{},[158,1754,1756],{"href":1755},"/blog/openclaw-not-working","OpenClaw Not Working: Every Fix in One Guide"," — Master troubleshooting guide for all common setup issues",[1302,1759,1760,1764],{},[158,1761,1763],{"href":1762},"/blog/openclaw-oom-errors","OpenClaw OOM Errors: Complete Fix Guide"," — Memory crashes that can trigger restart loops",[1302,1766,1767,1771],{},[158,1768,1770],{"href":1769},"/blog/openclaw-memory-fix","OpenClaw Memory Fix Guide"," — Context compaction issues that cause agents to lose track mid-task",[1302,1773,1774,1777],{},[158,1775,1776],{"href":1497},"OpenClaw API Costs: What You'll Actually Pay"," — Understand the cost impact of runaway loops",{"title":143,"searchDepth":423,"depth":423,"links":1779},[1780,1781,1782,1787,1788,1789,1790,1791,1792,1793,1794],{"id":1407,"depth":423,"text":1408},{"id":1438,"depth":423,"text":1439},{"id":1466,"depth":423,"text":1467,"children":1783},[1784,1785,1786],{"id":1477,"depth":434,"text":1478},{"id":1487,"depth":434,"text":1488},{"id":1502,"depth":434,"text":1503},{"id":1515,"depth":423,"text":1516},{"id":1542,"depth":423,"text":1543},{"id":1588,"depth":423,"text":1589},{"id":1611,"depth":423,"text":1612},{"id":1637,"depth":423,"text":1638},{"id":1674,"depth":423,"text":1675},{"id":348,"depth":423,"text":349},{"id":1296,"depth":423,"text":1297},"2026-03-26","OpenClaw agent stuck in a loop and burning API tokens? Learn why agents loop, what it costs, and how to add guardrails that stop the bleeding fast.","/img/blog/openclaw-agent-stuck-in-loop.jpg",{},"12 min read",{"title":1365,"description":1796},"OpenClaw Agent Stuck in Loop? Stop Burning $25+/Min","blog/openclaw-agent-stuck-in-loop",[1804,1805,1806,1807,1808,1809],"OpenClaw agent stuck in loop","OpenClaw loop fix","AI agent runaway cost","OpenClaw retry storm","OpenClaw circuit breaker","managed OpenClaw hosting","2026-04-02","r8vf1SNdzUrPLKz4gX-cSo2FdOpSo9HbMIB8z5puqEY",1779278332646]