MCP Helm Server

by zekker6

26 370 downloads Not rated yet MIT
GitHub

About

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…

Details

License
MIT

Explore

The MCP Helm server provides the following tools:

- list_repository_charts - Lists all charts available in a Helm repository (or chart name for OCI registries)
- list_chart_versions - Lists all available versions/tags for a chart
- get_latest_version_of_chart - Retrieves the latest version of a specific chart
- get_chart_values - Retrieves the values file for a chart (latest version or specific version)
- get_chart_contents - Retrieves the contents of a chart (including templates, values, and metadata)
- get_chart_dependencies - Retrieves the dependencies of a chart as defined in its Chart.yaml file
- get_chart_images - Extracts container images used in a Helm chart by rendering templates and parsing Kubernetes
manifests

Setting up with Highlight

This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:

  1. Download and install Highlight from highlightai.com/download
  2. Navigate to the plugins tab and select "Add Custom Plugin"
  3. Configure the plugin with the settings below
    Plugin Name MCP Helm Server
    Command (node, npx, python, etc.)

    Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.

  4. Enable "Start Automatically" if you want the plugin to start when Highlight launches

From the repository

Configure your MCP client to connect to this server. The server implements the standard MCP protocol for tool discovery
and execution.

There is a publicly available instance of the MCP Helm server that you can use to test the features without installing
it: https://mcp-helm.zekker.dev/mcp

> Note: Go 1.24.3 is required.

go install github.com/zekker6/mcp-helm/cmd/mcp-helm@latest

The server supports authentication for both OCI registries and HTTP Helm repositories.

For repositories requiring username/password authentication:


For private OCI registries, authentication can be configured via:

1. Docker credentials - The server automatically uses credentials from ~/.docker/config.json

2. Explicit credentials file - Use -registry-credentials flag
bash

docker login ghcr.io
echo $GITHUB_TOKEN | docker login ghcr.io -u USERNAME --password-stdin

./mcp-helm \
-username myuser -password-file /path/to/password.txt \
-registry-credentials /path/to/docker/config.json



Credentials behind an external credential store (credsStore) or per-registry

helper (credHelpers) are resolved by invoking that helper binary at runtime.

If the helper is not available in the runtime environment, the affected

registries fall back to basic auth; a warning is logged at startup so this is

visible. Routing considers only the file passed to -registry-credentials (no

implicit ~/.docker/config.json fallback), so list every private OCI registry

you need in that file.


For repositories with custom TLS requirements:
bash

./mcp-helm -tls-cert /path/to/client.crt -tls-key /path/to/client.key

Example with Docker, passing authentication:

```bash

Claude Desktop / Cursor

Paste into your MCP client config file to install this server.

{
    "mcpServers": {
        "mcp helm server": {
            "mcp-helm": {
                "command": "docker",
                "args": [
                    "run",
                    "-d",
                    "--name",
                    "mcp-helm",
                    "-p",
                    "8012:8012",
                    "ghcr.io/zekker6/mcp-helm:v1.3.0",
                    "-mode=sse"
                ]
            }
        }
    }
}

McpServers

{
    "mcp-helm": {
        "command": "docker",
        "args": [
            "run",
            "-d",
            "--name",
            "mcp-helm",
            "-p",
            "8012:8012",
            "ghcr.io/zekker6/mcp-helm:v1.3.0",
            "-mode=sse"
        ]
    }
}

An MCP (Model Context Protocol) server that provides tools for interacting with Helm repositories and charts. This
server enables AI assistants to query Helm repositories, retrieve chart information, and access chart values without
requiring local Helm installation.

The purpose of using MCP for Helm is to avoid making up format of values.yaml and contents of the charts when working
with LLMs.
Instead, the server provides a standardized way to access this information, making it easier for AI assistants to
interact with Helm charts and repositories.

This MCP server is and will be providing tools for working with Helm repositories only. If you need to work with other
Kubernetes resources, consider using a separate MCP server that provides tools for Kubernetes resources.

Features

The MCP Helm server provides the following tools:

- list_repository_charts - Lists all charts available in a Helm repository (or chart name for OCI registries)
- list_chart_versions - Lists all available versions/tags for a chart
- get_latest_version_of_chart - Retrieves the latest version of a specific chart
- get_chart_values - Retrieves the values file for a chart (latest version or specific version)
- get_chart_contents - Retrieves the contents of a chart (including templates, values, and metadata)
- get_chart_dependencies - Retrieves the dependencies of a chart as defined in its Chart.yaml file
- get_chart_images - Extracts container images used in a Helm chart by rendering templates and parsing Kubernetes
manifests

Repository Types

All tools support both traditional HTTP Helm repositories and OCI registries:

| Repository Type | Example URL |
|------------------|------------------------------------|
| HTTP Repository | https://charts.example.com |
| OCI Registry | oci://ghcr.io/org/charts/mychart |
| OCI (Docker Hub) | oci://docker.io/library/mysql |

OCI Registry Support

OCI (Open Container Initiative) registries store Helm charts as OCI artifacts. Unlike HTTP repositories where multiple
charts share an index, OCI registries typically contain one chart per repository with multiple version tags.

Example usage with OCI:

repository_url: oci://ghcr.io/nginxinc/charts/nginx-ingress
chart_name: (empty - chart name is in the URL)

Try without installation

There is a publicly available instance of the MCP Helm server that you can use to test the features without installing
it: https://mcp-helm.zekker.dev/mcp

Installation

Run with docker

You can run the MCP Helm server using Docker. This is the easiest way to get started without needing to install Go or
build from source.

docker run -d --name mcp-helm -p 8012:8012 ghcr.io/zekker6/mcp-helm:v1.3.0 -mode=sse

Note that the --mode=sse flag is used to enable Server-Sent Events mode, which used by MCP clients to connect.
Alternatively, you can use -mode=http to enable Streamable HTTP mode.

Via pre-build binary

Download binary from the releases page.

Example for Linux x86_64 (note that other architectures and platforms are also available):

latest=$(curl -s https://api.github.com/repos/zekker6/mcp-helm/releases/latest | grep 'tag_name' | cut -d\" -f4)
wget https://github.com/zekker6/mcp-helm/releases/download/$latest/mcp-helm_Linux_x86_64.tar.gz
tar axvf mcp-helm_Linux_x86_64.tar.gz

Via Mise

Mise (mise-en-place) is a development environment setup tool.

mise i ubi:zekker6/mcp-helm@latest

Install with Go

> Note: Go 1.24.3 is required.

go install github.com/zekker6/mcp-helm/cmd/mcp-helm@latest

Build from Source

> Note: Go 1.24.3 is required.

1. Clone the repository:

   git clone https://github.com/zekker6/mcp-helm.git

   cd mcp-helm

2. Build the binary:

   go build -o mcp-helm ./cmd/mcp-helm

3. Run the server:

   ./mcp-helm

Configuration

Configure your MCP client to connect to this server. The server implements the standard MCP protocol for tool discovery
and execution.

Authentication

The server supports authentication for both OCI registries and HTTP Helm repositories.

Command-Line Flags

| Flag | Description |
|-----------------------------|-----------------------------------------------------------------------|
| -username | Username for basic authentication (HTTP repos, and OCI registries not covered by -registry-credentials) |
| -password-file | Path to file containing password |
| -registry-credentials | Path to Docker-style credentials file (e.g., ~/.docker/config.json); authoritative for the OCI registries it lists |
| -registry-plain-http | Use plain HTTP for OCI registries (insecure, for development only) |
| -tls-cert | Path to TLS client certificate file for HTTP repositories |
| -tls-key | Path to TLS client key file for HTTP repositories |
| -tls-ca | Path to CA certificate file for verifying server certificates |
| -tls-insecure-skip-verify | Skip TLS certificate verification (insecure) |
| -pass-credentials-all | Pass credentials to all domains when following redirects |

Basic Authentication

For repositories requiring username/password authentication:

```bash

No reviews yet — be the first

Sign in to leave a review

Use Google, GitHub, or an email account so ratings stay tied to real people.

Email sign in

No reviews posted yet.

More like this

Similar MCP servers you might find useful

Browse category

Sequential Thinking

Sequential Thinking is an MCP server that provides a tool for dynamic and reflective problem-solving through a structured, step-by-step thinking process. It is…

— 88.2k 508.3k dl
Other

Sentry

Sentry MCP Server connects compatible AI clients with Sentry error monitoring, allowing models and agents to inspect issues and event data through Model…

— 88.1k 4.3k dl
Developer Tools

Github

GitHub MCP Server connects compatible AI clients with GitHub repositories, issues, and pull requests, allowing models and agents to read and act on project…

— 88.1k 9.9k dl
Developer Tools

Filesystem

Filesystem is a Node.js server that implements the Model Context Protocol (MCP) for filesystem operations. It enables AI assistants to read, write, and manage…

— 88.1k 6.7k dl
Other