home / mcp / rundeck mcp server

Rundeck MCP Server

Provides MCP endpoints to list jobs, run jobs with options, and fetch execution status and logs from Rundeck.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "justynroberts-rundeck-mcp": {
      "command": "uvx",
      "args": [
        "rundeck-mcp",
        "--enable-write-tools"
      ],
      "env": {
        "RUNDECK_URL": "http://localhost:4440",
        "RUNDECK_API_TOKEN": "${input:rundeck-api-token}",
        "RUNDECK_API_VERSION": "44"
      }
    }
  }
}

The Rundeck MCP Server lets you interact with your Rundeck instance from MCP-enabled clients. It provides read and write endpoints to list jobs, run jobs with options, monitor executions, and fetch logs, so you can automate and script Rundeck actions from your preferred MCP clients.

How to use

You connect to the MCP server from your MCP client and use the available tools to interact with Rundeck. Start the local MCP server using the runtime you prefer, then enable the server in your client settings. With write access enabled, you can run jobs and modify Rundeck state; without it, you can only read data.

How to install

Prerequisites you need before starting:

asdf-vm installed
uv installed globally
A Rundeck API Token

How to use the MCP server locally

# Prerequisites (commands shown separately)
# 1) Install asdf plugins
asdf plugin add python
asdf plugin add nodejs https://github.com/asdf-vm/asdf-nodejs.git
asdf plugin add uv

# 2) Install tool versions
asdf install

# 3) Create and activate a virtual environment if needed, then sync dependencies for uv-based project
uv sync

# 4) Ensure uv is available globally
# Follow the official guidance to add uv to your PATH if necessary
```

```
# Running the MCP server locally (example runtime configuration)
# This starts the local MCP server using uv and points to the module that implements the Rundeck MCP API
{
  "rundeck-mcp": {
    "type": "stdio",
    "command": "uv",
    "args": [
      "run",
      "--directory",
      "/path/to/your/mcp-server-directory",
      "python",
      "-m",
      "rundeck_mcp",
      "--enable-write-tools"
    ],
    "env": {
      "RUNDECK_API_TOKEN": "${input:rundeck-api-token}",
      "RUNDECK_URL": "http://localhost:4440"
    }
  }
}

Security and configuration notes

By default, the MCP server runs in read-only mode. To perform actions that modify Rundeck (such as running jobs), start the server with the --enable-write-tools flag. Protect your Rundeck API token and limit its scope to the jobs you need to manage.

Environment variables you use with the MCP server include the Rundeck API token and the Rundeck URL. Treat the API token as a secret and restrict access to systems that need it.

Client configuration examples

Two common ways to run and connect to the MCP server are shown below. Use these as references in your own client configuration.

Notes on integrations

Cursor, VS Code, Claude Desktop, and other MCP-enabled clients can load the following stdio server configuration and environment settings to interact with your Rundeck instance. Ensure you supply your actual Rundeck API token and Rundeck URL in place of placeholders.

Available tools

list_jobs

Lists jobs in a project with optional filtering (read-only)

get_job

Retrieves job details including options and defaults (read-only)

list_executions

Lists executions with filtering by status or time (read-only)

get_execution

Retrieves execution status and details (read-only)

get_execution_output

Retrieves execution log output (read-only)

run_job

Executes a job with options (write-capable)