K8sGPT

New: K8sGPT MCP Server with AI Integration!Exciting news! K8sGPT now features a powerful Model Context Protocol (MCP) server for Claude and other AI assistants.Learn moreGiving KubernetesSuperpowers to EveryoneK8sGPT is an AI-powered tool that helps diagnose and fix Kubernetes issues with intelligent insights and automated troubleshooting.View on GitHubGet StartedPowerful FeaturesEverything you need to manage and troubleshoot your Kubernetes clustersAI-Powered AnalysisAdvanced AI algorithms analyze your cluster state and provide intelligent insights for troubleshooting.Data AnonymizationSensitive data is automatically anonymized before being sent to AI backends for analysis.Multiple AI ProvidersSupport for various AI providers including OpenAI, Azure, Google, and local models.Auto RemediationAutomatically apply suggested fixes to common Kubernetes issues, reducing manual intervention and speeding up recovery.Claude Desktop IntegrationSeamlessly integrate K8sGPT with Claude Desktop for enhanced Kubernetes managementNative CLI ExperienceAccess K8sGPT directly through Claude Desktop's command line interface for a seamless development experience.Enhanced AI CapabilitiesLeverage Claude's advanced AI capabilities to get deeper insights into your Kubernetes clusters and workloads.Streamlined WorkflowCombine the power of K8sGPT's analysis with Claude's natural language understanding for a more intuitive troubleshooting experience.Learn more about MCP →Model Context Protocol (MCP)AI-Native Kubernetes IntegrationK8sGPT MCP server exposes powerful Kubernetes operations as standardized tools, resources, and prompts for any AI assistant.Cluster AnalysisAnalyze Kubernetes resources for issues and get AI-powered explanationsResource ManagementList and inspect any Kubernetes resource type with detailed informationLog & Events AccessRetrieve pod logs and cluster events for troubleshootingAnalyzer ControlManage and configure which analyzers to enable or disableWhat You Can Do12+ Powerful ToolsAccess cluster analysis, resource management, logs, events, and analyzer configuration through a standardized interface.Multiple Integration ModesChoose between Stdio mode for local AI assistants or HTTP mode for network access and webhooks.Rich Resources & PromptsAccess read-only cluster information and guided troubleshooting workflows for pods, deployments, and clusters.Standardized ProtocolBuilt on the Model Context Protocol (MCP) standard, enabling compatibility with any MCP-compatible AI client.Stdio ModePerfect for local AI assistants like Claude Desktopk8sgpt serve --mcpHTTP ModeEnable network access and webhook integrationsk8sgpt serve --mcp --mcp-http --mcp-port 8089Ready to supercharge your Kubernetes troubleshooting with AI?Explore MCP DocumentationView Integration GuideSupported AI ProvidersChoose from a wide range of AI providers to power your Kubernetes analysisOpenAIAzure OpenAIGoogle Vertex AIAmazon BedrockIBM WatsonXLocal AIOllamaCohereHugging FaceTrusted by Industry LeadersK8sGPT is trusted by industry leaders like Spectro Cloud, Nethopper, Upstage AI, and more within our community.Spectro CloudNethopperUpstage AIFine-Grained Control & GuardrailsK8sGPT puts you in control, allowing you to tailor analysis and remediation to your needs.Toggle Auto-Remediation: You decide if K8sGPT should automatically apply suggested fixes.AI-Free Analysis: Run built-in analyzers locally without involving any AI backend for simple diagnostics.Selective Analyzers: Choose precisely which analyzers to run, focusing only on the areas you need.Data Anonymization: Sensitive data is automatically anonymized before being sent to external AI backends.Local AI Models: Keep your data entirely within your environment by using supported local AI models.Join Our CommunityConnect with other K8sGPT users and contributorsCommunity MembershipJoin our growing community of users and contributors. We welcome everyone interested in Kubernetes and AI.Learn about membership →Office HoursJoin our regular office hours to get help, share your experiences, and connect with the K8sGPT team.Check schedule →Join the ConversationConnect with the community and maintainers on Slack for real-time discussions and support.Join Slack →K8sGPT in ActionSee how the community is using K8sGPT to enhance their Kubernetes workflows.Kubecon 2025 TalkLatest K8sGPT presentation from KubeCon --- New: K8sGPT MCP Server with AI Integration!Exciting news! K8sGPT now features a powerful Model Context Protocol (MCP) server for Claude and other AI assistants.Learn moreBack to HomeDocumentationDocumentation NoticeFor the most up-to-date and comprehensive documentation about K8sGPT, please visit:docs.k8sgpt.aiGitHub RepositoryGetting StartedOverviewK8sGPT gives Kubernetes SRE superpowers to everyone. Get started quickly with our installation guides and basic usage instructions.InstallationQuick start guide for installing K8sGPT on your system.Learn more →In-Cluster OperatorDeploy K8sGPT as an operator in your Kubernetes cluster.Learn more →CLI ConfigurationConfigure K8sGPT CLI settings and options.Learn more →Operator ConfigurationConfigure the K8sGPT Operator for your Kubernetes cluster.Learn more →TutorialsK8sGPT PlaygroundTry out K8sGPT features in an interactive environment.Start learning →Custom AnalyzersLearn how to create your own custom analyzers.Start learning →Slack IntegrationSet up K8sGPT notifications in Slack.Start learning →Custom Rest BackendIntegrate a custom backend with k8sgpt using RESTful API.Start learning →ObservabilityIntegrate Prometheus and Grafana with K8sGPT.Start learning →ReferenceCLI ReferenceOverview & InstallationBasic CommandsAdditional CommandsServe ModeAdditional ReferencesAI ProvidersOperator GuideMCP ServerExplanationIntegrationLearn about integrating K8sGPT with your existing Kubernetes infrastructure and other tools in your stack.Read more →CachingUnderstanding how K8sGPT uses caching to improve performance and reduce API calls to AI providers.Read more → --- New: K8sGPT MCP Server with AI Integration!Exciting news! K8sGPT now features a powerful Model Context Protocol (MCP) server for Claude and other AI assistants.Learn moreBack to DocumentationInstallationInstallation MethodsLearn how to install K8sGPT using various methods including CLI, operator, and package managers. CLI Installation The simplest way to get started with K8sGPT is using the CLI installation method. # Using curl curl -LO https://github.com/k8sgpt-ai/k8sgpt/releases/latest/download/k8sgpt_Linux_x86_64.tar.gz tar -xzf k8sgpt_Linux_x86_64.tar.gz sudo mv k8sgpt /usr/local/bin/ # Using wget wget https://github.com/k8sgpt-ai/k8sgpt/releases/latest/download/k8sgpt_Linux_x86_64.tar.gz tar -xzf k8sgpt_Linux_x86_64.tar.gz sudo mv k8sgpt /usr/local/bin/ Operator Installation For Kubernetes-native installation, use the K8sGPT operator. # Install the operator kubectl apply -f https://raw.githubusercontent.com/k8sgpt-ai/k8sgpt-operator/main/deploy/crds/k8sgpt.ai_v1alpha1_k8sgpt_cr.yaml kubectl apply -f https://raw.githubusercontent.com/k8sgpt-ai/k8sgpt-operator/main/deploy/operator.yaml # Create a K8sGPT instance kubectl apply -f https://raw.githubusercontent.com/k8sgpt-ai/k8sgpt-operator/main/deploy/crds/k8sgpt.ai_v1alpha1_k8sgpt_cr.yaml Package Manager Installation Homebrew (macOS) brew tap k8sgpt-ai/k8sgpt brew install k8sgpt APT (Debian/Ubuntu) # Add the repository curl -fsSL https://download.opensuse.org/repositories/home:/k8sgpt-ai:/stable/xUbuntu_22.04/Release.key | sudo gpg --dearmor -o /etc/apt/trusted.gpg.d/k8sgpt.gpg echo "deb https://download.opensuse.org/repositories/home:/k8sgpt-ai:/stable/xUbuntu_22.04/ /" | sudo tee /etc/apt/sources.list.d/k8sgpt.list # Install K8sGPT sudo apt update sudo apt install k8sgpt YUM/DNF (RHEL/CentOS) # Add the repository curl -fsSL https://download.opensuse.org/repositories/home:/k8sgpt-ai:/stable/RHEL_8/Release.key | sudo gpg --dearmor -o /etc/pki/rpm-gpg/RPM-GPG-KEY-k8sgpt echo "[k8sgpt] name=K8sGPT Repository baseurl=https://download.opensuse.org/repositories/home:/k8sgpt-ai:/stable/RHEL_8/ enabled=1 gpgcheck=1 gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-k8sgpt" | sudo tee /etc/yum.repos.d/k8sgpt.repo # Install K8sGPT sudo dnf install k8sgpt Docker Installation You can also run K8sGPT using Docker. docker run -it --rm -v ~/.kube:/root/.kube k8sgpt/k8sgpt:latest Verifying Installation After installation, verify that K8sGPT is working correctly: k8sgpt version Next Steps Now that you have K8sGPT installed, you can: Configure your AI backend Run your first analysis Set up auto-remediation Explore advanced features Check out our configuration guide to get started.← OverviewConfiguration → --- New: K8sGPT MCP Server with AI Integration!Exciting news! K8sGPT now features a powerful Model Context Protocol (MCP) server for Claude and other AI assistants.Learn moreBack to DocumentationModel Context Protocol (MCP) ServerK8sGPT MCP Server - Kubernetes operations as standardized tools for AI assistantsWhat is MCP?The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect to external data sources and tools. K8sGPT's MCP server exposes Kubernetes operations through this standardized interface, allowing AI assistants like Claude to:Analyze cluster health and issuesQuery Kubernetes resourcesAccess pod logs and eventsGet troubleshooting guidanceManage analyzer filtersQuick StartStart the MCP Server - Stdio ModeFor local AI assistants like Claude Desktop:k8sgpt serve --mcpStart the MCP Server - HTTP ModeFor network access and webhooks:k8sgpt serve --mcp --mcp-http --mcp-port 8089Server ModesStdio Mode (Default)Used by local AI assistants like Claude Desktop for direct process communication.Configuration Example:{ "mcpServers": { "k8sgpt": { "command": "k8sgpt", "args": ["serve", "--mcp"] } } }HTTP ModeUsed for network access and webhook integrations. The server runs in stateless mode with independent requests.Command:k8sgpt serve --mcp --mcp-http --mcp-port 8089Available ToolsThe MCP server exposes 12+ tools for Kubernetes operations:Cluster AnalysisanalyzeAnalyze Kubernetes resources for issues and problems with optional AI explanations.Parameters:namespace (optional): Namespace to analyzeexplain (optional): Get AI explanations for issuesfilters (optional): Comma-separated list of analyzers to usecluster-infoGet Kubernetes cluster information and version details.Resource Managementlist-resourcesList Kubernetes resources of a specific type.Parameters:resourceType (required): Type of resource (pods, deployments, services, nodes, jobs, cronjobs, statefulsets, daemonsets, replicasets, configmaps, secrets, ingresses, pvcs, pvs)namespace (optional): Namespace to querylabelSelector (optional): Label selector for filteringget-resourceGet detailed information about a specific Kubernetes resource.Parameters:resourceType (required): Type of resourcename (required): Resource namenamespace (optional): Namespacelist-namespacesList all namespaces in the cluster.Debugging and Troubleshootingget-logsGet logs from a pod container.Parameters:podName (required): Name of the podnamespace (optional): Namespacecontainer (optional): Container nametail (optional): Number of lines to showprevious (optional): Show logs from previous container instancesinceSeconds (optional): Show logs from last N secondslist-eventsList Kubernetes events for debugging.Parameters:namespace (optional): Namespace to queryinvolvedObjectName (optional): Filter by object nameinvolvedObjectKind (optional): Filter by object kindAnalyzer Managementlist-filtersList all available and active analyzers/filters.add-filtersAdd filters to enable specific analyzers.Parameters:filters (required): Comma-separated list of analyzer namesremove-filtersRemove filters to disable specific analyzers.Parameters:filters (required): Comma-separated list of analyzer namesIntegration & Configurationlist-integrationsList available integrations (Prometheus, AWS, Keda, Kyverno, etc.)configConfigure K8sGPT settings including custom analyzers and cache.Available ResourcesResources provide read-only access to cluster information:cluster-infoURI: cluster-info - Get information about the Kubernetes clusternamespacesURI: namespaces - List all namespaces in the clusteractive-filtersURI: active-filters - Get currently active analyzers/filtersAvailable PromptsPrompts provide guided troubleshooting workflows:troubleshoot-podInteractive pod debugging workflowArguments: podName (required), namespace (required)troubleshoot-deploymentInteractive deployment debugging workflowArguments: deploymentName (required), namespace (required)troubleshoot-clusterGeneral cluster troubleshooting workflowIntegration with AI AssistantsClaude DesktopAdd to claude_desktop_config.json:{ "mcpServers": { "k8sgpt": { "command": "k8sgpt", "args": ["serve", "--mcp"] } } }Restart Claude Desktop and you'll see k8sgpt tools available in the tool selector.Custom MCP ClientsAny MCP-compatible client can connect to the k8sgpt server:Start the server: k8sgpt serve --mcp --mcp-http --mcp-port 8089Connect to: http://localhost:8089/mcpUse standard MCP protocol methods: tools/list, tools/call, resources/read, prompts/getAdvanced ConfigurationCustom Portk8sgpt serve --mcp --mcp-http --mcp-port 9000With Specific Backendk8sgpt serve --mcp --backend openaiWith Kubeconfigk8sgpt serve --mcp --kubeconfig ~/.kube/configTroubleshootingConnection IssuesVerify the server is running:curl http://localhost:8089/mcpPermission IssuesEnsure your kubeconfig has appropriate cluster access:kubectl cluster-infoTool ErrorsList available tools to verify names:curl -X POST http://localhost:8089/mcp \ -H "Content-Type: application/json" \ -d '{"jsonrpc": "2.0", "id": 1, "method": "tools/list"}'Learn MoreMCP SpecificationK8sGPT DocumentationMCP Server GitHub DocumentationMCP Go Library← ProvidersIntegration →