talkopsai/argocd-mcp-server repository overview

⁠ArgoCD MCP Server

A comprehensive Model Context Protocol (MCP) server for managing Kubernetes workloads via ArgoCD using GitOps principles. This image allows AI assistants (like Claude) to perform secure, production-grade ArgoCD operations, including repository onboarding, application deployments, rollbacks, and intelligent debugging.

⁠✨ Features

• End-to-End Deployments: Create applications, preview deployment diffs, and orchestrate manual or auto-sync deployments.
• Smart Debugging: Analyze sync states, automatically extract and summarize raw pod error logs (e.g., CrashLoopBackOff, Exceptions), and fetch Kubernetes event histories.
• Repository Management: Securely onboard GitHub/GitLab repositories via HTTPS or SSH directly from chat.
• Multi-Tenant Projects: Create and manage ArgoCD projects to isolate teams, restrict namespaces, and control repository access.
• Safe Rollbacks: Intelligently rollback deployments to previous states or exact Git commit hashes.
• Declarative GitOps: Generate infrastructure-as-code manifests (like AppProject or v1/Secret) to commit back to Git.
• Built-in Guidance: Includes built-in workflow prompts to guide the AI step-by-step through complex deployment and debugging scenarios.

⁠🚀 How to use this image

This Docker image is designed to be used as an MCP server. It requires access to your ArgoCD API server and authentication credentials.

⁠Running Standalone (HTTP Transport)

Run the container mapping port 8770 so clients can connect locally:

docker run --rm -it \
  -p 8770:8770 \
  -e ARGOCD_SERVER_URL="https://host.docker.internal:8080" \
  -e ARGOCD_AUTH_TOKEN="your-token-here" \
  -e MCP_PORT=8770 \
  -e MCP_PATH="/mcp" \
  -e ARGOCD_INSECURE="true" \
  -e GIT_PASSWORD="your-github-pat" \
  -e MCP_ALLOW_WRITE=true \
  -e GIT_USERNAME="your-github-username" \
  talkopsai/argocd-mcp-server:latest

Then configure your MCP client to connect over HTTP/SSE:

{
  "mcpServers": {
    "argocd-mcp-server": {
      "serverUrl": "http://localhost:8770/mcp",
      "disabled": false,
      "disabledTools": []
    }
  }
}

⁠Important: Environment Variables

Authentication is entirely handled via environment variables to keep your credentials secure. At a minimum, you must provide ARGOCD_SERVER_URL and ARGOCD_AUTH_TOKEN. If you plan on having the AI onboard private git repositories, you should also provide GIT_USERNAME and GIT_PASSWORD.

⁠🛠️ Configuration (Claude Desktop - Stdio)

To use this image securely over stdio transport directly inside Claude Desktop, format the invocation like this:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "argocd-mcp-server": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e", "ARGOCD_SERVER_URL=https://host.docker.internal:8080",
        "-e", "ARGOCD_AUTH_TOKEN=your_argocd_token",
        "-e", "ARGOCD_INSECURE=true",
        "-e", "GIT_USERNAME=your_git_username",
        "-e", "GIT_PASSWORD=YOUR_GITHUB_PAT",
        "-e", "MCP_ALLOW_WRITE=true",
        "-e", "MCP_LOG_LEVEL=INFO",
        "-e", "MCP_TRANSPORT=stdio",
        "talkopsai/argocd-mcp-server:latest"
      ]
    }
  }
}

⁠⚙️ Environment Variables

The image supports several environment variables for authentication and advanced configuration:

⁠🔐 Security Best Practices

1. Personal Access Tokens: Always use a Personal Access Token (PAT) for the GIT_PASSWORD environment variable rather than your actual Git account password.
2. ArgoCD Least Privilege: Ensure the ARGOCD_AUTH_TOKEN belongs to an ArgoCD Service Account or Role that is strictly scoped (via ArgoCD RBAC) to the projects and permissions the AI needs. Do not provide a cluster-admin token if not necessary.
3. Hidden Credentials: Never expose your passwords in chat. By pre-loading them into the Docker container via environment variables, the AI can authenticate infrastructure securely under the hood without ever logging the secrets.

⁠🔗 Links and Resources

• Source Code: GitHub Repository⁠
• Model Context Protocol: MCP Official Site⁠
• ArgoCD: ArgoCD Official Site⁠
• Report Issues: GitHub Issues⁠

If you find this MCP server useful, consider leaving a ⭐ on the GitHub repository⁠!