rundeck-mcp-server
MCP server enabling AI assistants to manage Rundeck projects, jobs, and executions through a standardized interface.
README
Rundeck MCP Server
A Model Context Protocol (MCP) server for Rundeck, enabling AI assistants to manage projects, jobs, and executions through a standardized interface.
Prerequisites
- Python 3.12+
- uv package manager
- A Rundeck instance with API access
- A Rundeck API Token (generate under User Profile > API Tokens in the Rundeck UI)
Quick Start
Install and run directly with uvx (no clone needed):
uvx rundeck-mcp --enable-write-tools
Or install from GitHub:
uv pip install git+https://github.com/justynroberts/rundeck-mcp-server.git
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
RUNDECK_API_TOKEN |
Yes | - | Rundeck API authentication token |
RUNDECK_URL |
No | http://localhost:4440 |
Rundeck server base URL |
RUNDECK_API_VERSION |
No | 41 |
Rundeck API version |
MCP Client Configuration
Claude Code
From a local clone:
git clone https://github.com/justynroberts/rundeck-mcp-server.git
claude mcp add rundeck \
-e RUNDECK_API_TOKEN=<your-api-token> \
-e RUNDECK_URL=http://your-rundeck-server:4440 \
-- uv run --directory /path/to/rundeck-mcp-server rundeck-mcp --enable-write-tools
Or add manually to ~/.claude/settings.json:
{
"mcpServers": {
"rundeck": {
"command": "uv",
"args": [
"run", "--directory", "/path/to/rundeck-mcp-server",
"rundeck-mcp", "--enable-write-tools"
],
"env": {
"RUNDECK_API_TOKEN": "<your-api-token>",
"RUNDECK_URL": "http://your-rundeck-server:4440"
}
}
}
}
From PyPI (once published):
claude mcp add rundeck \
-e RUNDECK_API_TOKEN=<your-api-token> \
-e RUNDECK_URL=http://your-rundeck-server:4440 \
-- uvx rundeck-mcp --enable-write-tools
Claude Desktop
Add to your Claude Desktop MCP configuration:
{
"mcpServers": {
"rundeck": {
"command": "uvx",
"args": ["rundeck-mcp", "--enable-write-tools"],
"env": {
"RUNDECK_API_TOKEN": "<your-api-token>",
"RUNDECK_URL": "http://your-rundeck-server:4440"
}
}
}
}
Cursor
Add to .cursor/mcp.json:
{
"mcpServers": {
"rundeck": {
"command": "uvx",
"args": ["rundeck-mcp", "--enable-write-tools"],
"env": {
"RUNDECK_API_TOKEN": "<your-api-token>",
"RUNDECK_URL": "http://your-rundeck-server:4440"
}
}
}
}
VS Code
Add to .vscode/mcp.json:
{
"servers": {
"rundeck": {
"command": "uvx",
"args": ["rundeck-mcp", "--enable-write-tools"],
"env": {
"RUNDECK_API_TOKEN": "<your-api-token>",
"RUNDECK_URL": "http://your-rundeck-server:4440"
}
}
}
}
Local Development
git clone https://github.com/justynroberts/rundeck-mcp-server.git
cd rundeck-mcp-server
uv sync --group dev
Run tests
uv run pytest
Lint and type check
uv run ruff check .
uv run pyright
Run locally
export RUNDECK_API_TOKEN=<your-token>
export RUNDECK_URL=http://localhost:4440
uv run rundeck-mcp --enable-write-tools
Available Tools
Read Tools (always available)
| Tool | Description |
|---|---|
list_projects |
List all Rundeck projects |
get_project |
Get details for a specific project |
list_jobs |
List jobs in a project with optional filtering |
get_job |
Get full job definition including options and steps |
get_execution |
Get current status of a job execution |
list_executions_for_job |
List recent executions for a job |
get_execution_output |
Get log output from an execution |
Write Tools (require --enable-write-tools)
| Tool | Description |
|---|---|
create_project |
Create a new Rundeck project |
import_job_yaml |
Import a job from YAML definition |
delete_job |
Delete a job |
run_job |
Execute a job with optional parameters |
abort_execution |
Abort a running execution |
Job Execution Workflow
The server enforces a safe execution pattern:
- Discover: Use
get_jobto retrieve the job definition and its options - Present: The AI presents option names, descriptions, defaults, required flags, and allowed values to the user
- Collect: The user provides values for required and desired optional parameters
- Execute:
run_jobis called with the collected option values - Monitor: Use
get_executionandget_execution_outputto track progress
License
MIT License - Copyright (c) fintonlabs.com
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。