Helm MCP

mcp-helm

[](https://github.com/kubedoll-heavy-industries/helm-mcp/actions/workflows/ci.yml) [](https://codecov.io/gh/kubedoll-heavy-industries/helm-mcp) [](https://goreportcard.com/report/github.com/kubedoll-heavy-industries/helm-mcp) [](https://github.com/kubedoll-heavy-industries/helm-mcp/releases/latest) [](LICENSE)

Give your AI assistant access to real Helm chart data. No more hallucinated values.yaml files.

What is this?

When you ask Claude, Cursor, or other AI assistants to help with Kubernetes deployments, they don't have access to Helm chart schemas. So they guess — and the guesses look plausible but don't match reality.

Without mcp-helm:

• :x: Hallucinates field names that look right but don't exist
• :x: Suggests stale or deprecated chart versions
• :x: Wastes tokens on web fetches and guesswork

With mcp-helm:

• :whitecheckmark: Queries actual Helm repositories for real chart data
• :whitecheckmark: Gets the latest chart version automatically
• :whitecheckmark: Correct configurations the first time

mcp-helm implements the Model Context Protocol (MCP) — a standard way for AI assistants to access external data sources.

Try It Now

Add this to your editor's MCP config to use our public instance (rate limited, no install required):

{
  "mcpServers": {
    "helm": {
      "type": "http",
      "url": "https://helm-mcp.kubedoll.com/mcp"
    }
  }
}

Then ask your AI: "What values can I configure for the bitnami/postgresql chart?"

Editor Setup

Claude Code

Edit ~/.claude/mcp.json:

{
  "mcpServers": {
    "helm": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "ghcr.io/kubedoll-heavy-industries/mcp-helm", "--transport=stdio"]
    }
  }
}

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "helm": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "ghcr.io/kubedoll-heavy-industries/mcp-helm", "--transport=stdio"]
    }
  }
}

Cursor

Edit MCP settings in Cursor's configuration:

{
  "mcpServers": {
    "helm": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "ghcr.io/kubedoll-heavy-industries/mcp-helm", "--transport=stdio"]
    }
  }
}

VS Code + Continue

Add to your Continue config (~/.continue/config.json):

{
  "experimental": {
    "modelContextProtocolServers": [
      {
        "transport": {
          "type": "stdio",
          "command": "docker",
          "args": ["run", "--rm", "-i", "ghcr.io/kubedoll-heavy-industries/mcp-helm", "--transport=stdio"]
        }
      }
    ]
  }
}

Without Docker

If you prefer to run the binary directly, [install mcp-helm](#install) and replace the Docker config with:

{
  "mcpServers": {
    "helm": {
      "command": "mcp-helm"
    }
  }
}

Available Tools

| Tool | What it does | Useful parameters | |------|--------------|-------------------| | search_charts | List or search charts in a Helm repo | keyword (substring filter), limit | | get_versions | Get available versions of a chart (newest first) | limit=1 for the latest only | | get_values | Get chart values.yaml, optionally as a focused subsection | path (e.g. .ingress), depth (default 2, 0 for full YAML), include_schema=true, include_examples=true (requires path) | | get_dependencies | Get a chart's sub-charts (with their repo URLs, which can be fed back into the other tools) | — | | get_notes | Get chart NOTES.txt (post-install instructions) | — |

OCI registries (oci://...) do not support browsing — for OCI you must already know the chart name, then call get_versions or get_values directly with that name.

Install

Docker (recommended — no install required, used in Editor Setup above):

docker pull ghcr.io/kubedoll-heavy-industries/mcp-helm:latest

Binary:

curl -fsSL https://github.com/kubedoll-heavy-industries/helm-mcp/releases/latest/download/mcp-helm_$(uname -s)_$(uname -m).tar.gz | tar xz
sudo mv mcp-helm /usr/local/bin/

Go:

go install github.com/kubedoll-heavy-industries/helm-mcp/cmd/mcp-helm@latest

Self-Hosting

For shared deployments or when you need an HTTP endpoint:

docker run -p 8012:8012 ghcr.io/kubedoll-heavy-industries/mcp-helm:latest \
  --transport=http --listen=:8012
# Connect to http://localhost:8012/mcp

See [docs/self-hosting.md](docs/self-hosting.md) for health endpoints and production recommendations.

Documentation

• [Configuration Reference](docs/configuration.md) — CLI flags, env vars, transport modes
• [Self-Hosting Guide](docs/self-hosting.md) — Docker HTTP, health endpoints, production tips
• [Troubleshooting](docs/troubleshooting.md) — common issues and fixes
• [Contributing](docs/contributing.md) — development setup, testing, PR guidelines
• [Security Policy](SECURITY.md) — reporting vulnerabilities

License

MIT — see [LICENSE](LICENSE).

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: kubedoll-heavy-industries
• Source: kubedoll-heavy-industries/helm-mcp
• License: MIT
• Homepage: https://helm-mcp.kubedoll.com

Install and usage instructions live in the source repository linked above.