mcp-jenkins

This MCP server integrates Jenkins with AI assistants by exposing Jenkins operations through the MCP (Model Context Protocol) interface. It acts as a bridge that lets you securely query Jenkins items, builds, nodes, queues, and more from AI tooling. The server supports different transport modes, including streamable-http, SSE, and stdio, allowing you to select the interaction style that best fits your setup. Available tools cover a broad range of Jenkins activities, such as retrieving items, builds, and configurations, or triggering and managing builds and queue items. This makes it possible to query and control your Jenkins environment contextually while preserving separation between the AI workspace and CI/CD infrastructure.

To use it, install the MCP server (via uvx as shown) and run it with the Jenkins endpoint details you need. The most common workflow is to run the server with streamable-http or SSE transport for robust, streaming interactions, and then configure your IDE or Copilot integration to connect to the provided MCP endpoint (e.g., http://localhost:9887/mcp or the SSE endpoint at http://localhost:9887/sse). The toolset includes actions like get_item, get_all_items, build_item, get_build, get_node, and many others to inspect and trigger Jenkins activities in a structured, AI-friendly way.

Prerequisites:

• Python 3.8+ (or an environment where uv is available)
• pip installed with Python
• Optional: Docker for containerized usage

Install via uv (recommended):

pip install uv
uvx mcp-jenkins

Install directly with pip:

pip install mcp-jenkins
mcp-jenkins

Docker (alternative):

docker pull ghcr.io/lanbaoshen/mcp-jenkins:latest
docker run -p 9887:9887 --rm ghcr.io/lanbaoshen/mcp-jenkins:latest --transport streamable-http

Usage note: You can pass Jenkins connection details via command line when launching, for example:

uvx mcp-jenkins --jenkins-url http://your-jenkins-host --jenkins-username your-username --jenkins-password your-password --transport sse

Tips and considerations:

• Environment variables: you may expose Jenkins credentials via headers when using HTTP transports; the README lists supported flags for Jenkins URL, username, password, timeout, SSL verification, and session singleton.
• Transport choices: streamable-http and SSE are common for modern IDE integrations; stdio is useful for local debugging and direct integration in the terminal.
• If you plan to expose the MCP endpoint over a network, ensure proper authentication and limit access to trusted clients.
• The toolset includes a broad range of Jenkins operations (get_item, get_all_items, build_item, get_build, stop_build, etc.). Review the Available Tools table in the README to understand each capability.
• For JetBrains Copilot or VSCode Copilot configurations, follow the JSON examples in the Configuration and Usage section to wire the MCP server into your editor workflow.