What is MCP Helm Server?
The MCP Helm Server is a Model Context Protocol server that provides tools for AI assistants to query Helm repositories and charts without requiring a local Helm installation. Its purpose is to avoid LLMs making up the format of values.yaml and chart contents, offering a standardized way to access this information.
How to use MCP Helm Server?
Install it via Docker, a pre-built binary from the releases page, Mise, go install, or building from source. Then configure your MCP client to connect using the server’s SSE mode (e.g., -mode=sse) or Streamable HTTP mode. Authentication credentials for private repositories can be passed with command-line flags like -username, -password-file, and -registry-credentials.
Key features of MCP Helm Server
- List all charts in a Helm repository (
list_repository_charts)
- List all available versions for a chart (
list_chart_versions)
- Get the latest version of a chart (
get_latest_version_of_chart)
- Retrieve a chart’s values file (
get_chart_values)
- Retrieve full chart contents including templates and metadata (
get_chart_contents)
- Extract chart dependencies from
Chart.yaml (get_chart_dependencies)
- Extract container images used in a chart (
get_chart_images)
- Supports both HTTP Helm repositories and OCI registries
Use cases of MCP Helm Server
- An AI assistant queries which versions of a Helm chart are available in a repository
- A developer retrieves the
values.yaml file for a specific chart version without local Helm
- A CI pipeline checks for updated chart dependencies before deploying
- An LLM extracts container images from a chart to verify or document resource usage
コメント