rootcause

Overview

RootCause is built in Go as a single binary that provides AI-powered SRE tools for Kubernetes clusters. It offers evidence-backed incident analysis, Kubernetes diagnostics, and safer operations with comprehensive guardrails including role/namespace policy enforcement, redaction, read-only mode, and destructive tool controls. The server exposes extensive tools for Kubernetes operations, ecosystem diagnostics (ArgoCD, Flux, cert-manager, etc.), Helm and Terraform workflows, and cloud context integration.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• rootcause.incident_bundle

  Collects evidence for Kubernetes incidents across multiple components

  Try:Generate an incident bundle for namespace paymentsTry:Collect evidence for the failed deployment
• rootcause.rca_generate

  Generates root cause analysis for incidents

  Try:Analyze the root cause of this deployment failureTry:Find why this pod keeps crashing
• rootcause.change_timeline

  Tracks changes in the Kubernetes environment

  Try:Show what changed before the incidentTry:Display the timeline of recent deployments
• k8s.restart_safety_check

  Checks if it's safe to restart a workload

  Try:Is it safe to restart the payment-api deployment?Try:Check if I can restart this pod without issues
• k8s.best_practice

  Runs Kubernetes best practice checks on deployments

  Try:Check the payment-api deployment for best practicesTry:Run best practice checks on all deployments in this namespace
• k8s.safe_mutation_preflight

  Validates safety before Kubernetes mutations

  Try:Check if this apply operation is safe to runTry:Validate safety before patching this deployment
• helm.chart_search

  Searches for Helm charts

  Try:Find Helm charts for databaseTry:Search for Prometheus chart
• helm.release_diff

  Compares Helm release versions

  Try:Show differences between current and desired releaseTry:Diff my current payment-api release
• terraform.debug_plan

  Analyzes Terraform plans for potential issues

  Try:Debug this Terraform planTry:Check risks in my infrastructure changes
• argocd.detect

  Detects ArgoCD installation

  Try:Check if ArgoCD is installed in this clusterTry:Detect ArgoCD presence
• flux.detect

  Detects Flux installation

  Try:Check if Flux is installed in this clusterTry:Detect Flux controllers
• k8s.diagnose_workload

  Diagnoses issues with Kubernetes workloads

  Try:Diagnose issues with the payment-api deploymentTry:Find why this pod is in a crash loop

Comparable tools

Installation

chmod +x rootcause

sudo mv rootcause /usr/local/bin/

rootcause --version

{
  "mcpServers": {
    "rootcause": {
      "command": "rootcause",
      "args": ["mcp"]
    }
  }
}

FAQ

What makes RootCause different from other Kubernetes tools?

RootCause is specifically designed as an MCP server that translates natural language requests into evidence-backed incident analysis with built-in safety controls, rather than requiring users to learn specific CLI commands or APIs.

Does RootCase require cluster access to function?

Yes, RootCause uses your existing kubeconfig for authentication and requires access to the Kubernetes API to provide diagnostics and analysis. It works with your current kubeconfig identity.

On Hacker News

Recent discussion from the developer community.

• Kubernetes MCP Server▲ 1
  Story by yindia · 2026-02-07

Compare rootcause with