Evergreen MCP Server

[!WARNING] This project is deprecated. MongoDB users should use the DevProd MCP server instead, which provides Evergreen tools alongside other DevProd integrations. This repository is no longer the recommended way to connect AI assistants to Evergreen, and the setup instructions below are kept for reference only.

A Model Context Protocol (MCP) server that provides access to the Evergreen CI/CD platform API. This server enables AI assistants and other MCP clients to interact with Evergreen projects, builds, tasks, and other CI/CD resources.

Overview

Evergreen is MongoDB’s continuous integration platform. This MCP server exposes Evergreen’s functionality through the Model Context Protocol, allowing AI assistants to help with CI/CD operations, project management, and build analysis.

Features

• Project Resources: Access and list Evergreen projects and build statuses
• Failed Jobs Analysis: Fetch failed jobs and logs for specific commits to help identify CI/CD failures
• Unit Test Failure Analysis: Detailed analysis of individual unit test failures with test-specific logs and metadata
• Task Log Retrieval: Get detailed logs for failed tasks with error filtering
• REST API Log Analysis: Full untruncated task and test logs via REST API with automatic error pattern scanning
• Stepback Analysis: Find failed mainline tasks that have undergone stepback bisection
• Authentication: Secure OIDC-based authentication via evergreen login
• Async Operations: Built on asyncio for efficient concurrent operations
• GraphQL + REST Integration: Uses Evergreen’s GraphQL API for metadata and REST API for full log content

Quick Start

Step 1: Authenticate with Evergreen

First, authenticate with Evergreen using the CLI. This creates the necessary credentials that the MCP server will use:

evergreen login

This will:

• Open your browser for OIDC authentication
• Create ~/.evergreen.yml with your credentials
• Create ~/.kanopy/token-oidclogin.json with your OIDC token

Note: If you don’t have the Evergreen CLI installed, see Evergreen CLI Installation.

Step 2: Configure Your MCP Client

Add the Evergreen MCP server to your AI assistant’s MCP configuration. You can use either uv (lightweight, no Docker needed) or Docker.

Option A: Using uv (Recommended)

uv is a fast Python package manager that can run the MCP server directly — no cloning, no virtual environments, no Docker required.

Install uv (if you don’t have it):

curl -LsSf https://astral.sh/uv/install.sh | sh

Then add the server to your MCP client config: