--- title: Can An AI Agent Debug Kubernetes? I Broke A Cluster To... description: An engineer deliberately breaks a local Kubernetes cluster with five common failure modes (OOMKilled, CrashLoopBackOff, ImagePullBackOff, readiness probe... canonical: https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Can An AI Agent Debug Kubernetes? I Broke A Cluster To Find Out | daily.dev og:description: An engineer deliberately breaks a local Kubernetes cluster with five common failure modes (OOMKilled, CrashLoopBackOff, ImagePullBackOff, readiness probe... og:url: https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr og:image: https://api.daily.dev/og/posts/lpg5q9Gwr.png og:image:alt: Can An AI Agent Debug Kubernetes? I Broke A Cluster To Find Out og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Can An AI Agent Debug Kubernetes? I Broke A Cluster To Find Out **[marmelab](https://daily.dev/sources/Marmelab)** · 9 min read · 9 upvotes · 0 comments ## Summary An engineer deliberately breaks a local Kubernetes cluster with five common failure modes (OOMKilled, CrashLoopBackOff, ImagePullBackOff, readiness probe failure, and a scaled-down dependency) and tests three open-source AI agents against it: k8sgpt, kubectl-ai, and kagent. k8sgpt runs deterministic analyzers with an optional LLM explanation step and works well as an MCP server inside Claude Code, but doesn't explore the cluster autonomously. kubectl-ai is more agentic, exploring the cluster and proposing fixes through natural-language interaction, and becomes the author's preferred primary tool. kagent, deployed as in-cluster CRDs with a chat dashboard, successfully diagnoses and fixes issues with the best agentic experience but requires installing an agent inside the cluster, which the author is reluctant to do on client infrastructure. A bonus section covers Docker Desktop's Gordon agent doing the same job for Compose stacks. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the difference between k8sgpt and kubectl-ai for diagnosing Kubernetes issues? k8sgpt runs deterministic analyzers on a cluster and optionally sends the output to an LLM for a one-shot explanation via its --explain flag, without exploring the cluster interactively. kubectl-ai is more agentic: it lets the LLM call tools to explore the cluster itself, round-tripping until it can explain and propose a fix, and supports natural-language follow-up questions. _Weighing agentic Kubernetes debugging tools is easier when comparisons like this surface on daily.dev._ ### Can an AI agent actually fix a broken Kubernetes cluster automatically? Yes, kagent successfully diagnosed and repaired a deliberately broken cluster without any code being written, by having the agent propose fixes for issues like OOMKilled pods and CrashLoopBackOff, which the operator then approved before the agent applied them. kagent runs as in-cluster CRDs with a chat dashboard, though full autonomous remediation without human approval was not tested. _daily.dev helps engineers track how far autonomous cluster remediation has actually progressed._ ### What Kubernetes failure scenarios are useful for testing an AI debugging agent? A solid test set includes an OOMKilled pod (memory limit of 20Mi with 250M allocated), a CrashLoopBackOff pod that exits with an error code, an ImagePullBackOff caused by a bad image tag, a readiness probe pointed at a 404 path, and a dependent service like redis-cart scaled to zero replicas causing HTTP 500s across the app. _Engineers benchmarking AI ops tools can find similar hands-on experiments curated on daily.dev._ ## Similar posts on daily.dev - [Agentic Operations for Kubernetes: AI Agents Replacing Manual K8s Management](https://daily.dev/posts/agentic-operations-for-kubernetes-ai-agents-replacing-manual-k8s-management-unlr4pesj) · Cast AI · 1 upvotes · 0 comments - [AI-driven Kubernetes in Action: Exploring AI-Assisted Kubernetes Operations](https://daily.dev/posts/ai-driven-kubernetes-in-action-exploring-ai-assisted-kubernetes-operations-uo6kz7s2k) · Cloud Native Now · 1 upvotes · 0 comments --- Tags: [#devops](https://daily.dev/tags/devops), [#kubernetes](https://daily.dev/tags/kubernetes), [#ai-agents](https://daily.dev/tags/ai-agents) [View this post on daily.dev](https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Can An AI Agent Debug Kubernetes? I Broke A Cluster To Find Out","url":"https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr"},"datePublished":"2026-09-03T11:16:16.318Z","dateModified":"2026-09-03T11:27:34.087Z","description":"An engineer deliberately breaks a local Kubernetes cluster with five common failure modes (OOMKilled, CrashLoopBackOff, ImagePullBackOff, readiness probe...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/e432aff099bee27ea73362e974799719?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/e432aff099bee27ea73362e974799719?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"marmelab","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"marmelab","logo":"https://media.daily.dev/image/upload/s--AeMXoaI9--/f_auto/v1716621144/logos/Marmelab","url":"https://daily.dev/sources/Marmelab"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":9},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"devops,kubernetes,ai-agents","timeRequired":"PT9M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"marmelab","item":"https://daily.dev/sources/Marmelab"},{"@type":"ListItem","position":3,"name":"Can An AI Agent Debug Kubernetes? I Broke A Cluster To Find Out"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/can-an-ai-agent-debug-kubernetes-i-broke-a-cluster-to-find-out-lpg5q9gwr#faq","mainEntity":[{"@type":"Question","name":"What is the difference between k8sgpt and kubectl-ai for diagnosing Kubernetes issues?","acceptedAnswer":{"@type":"Answer","text":"k8sgpt runs deterministic analyzers on a cluster and optionally sends the output to an LLM for a one-shot explanation via its --explain flag, without exploring the cluster interactively. kubectl-ai is more agentic: it lets the LLM call tools to explore the cluster itself, round-tripping until it can explain and propose a fix, and supports natural-language follow-up questions. Weighing agentic Kubernetes debugging tools is easier when comparisons like this surface on daily.dev."}},{"@type":"Question","name":"Can an AI agent actually fix a broken Kubernetes cluster automatically?","acceptedAnswer":{"@type":"Answer","text":"Yes, kagent successfully diagnosed and repaired a deliberately broken cluster without any code being written, by having the agent propose fixes for issues like OOMKilled pods and CrashLoopBackOff, which the operator then approved before the agent applied them. kagent runs as in-cluster CRDs with a chat dashboard, though full autonomous remediation without human approval was not tested. daily.dev helps engineers track how far autonomous cluster remediation has actually progressed."}},{"@type":"Question","name":"What Kubernetes failure scenarios are useful for testing an AI debugging agent?","acceptedAnswer":{"@type":"Answer","text":"A solid test set includes an OOMKilled pod (memory limit of 20Mi with 250M allocated), a CrashLoopBackOff pod that exits with an error code, an ImagePullBackOff caused by a bad image tag, a readiness probe pointed at a 404 path, and a dependent service like redis-cart scaled to zero replicas causing HTTP 500s across the app. Engineers benchmarking AI ops tools can find similar hands-on experiments curated on daily.dev."}}]} ```