MCP Atlassian

A Model Context Protocol (MCP) server that provides comprehensive tools for interacting with Atlassian products (Confluence and Jira).

Overview

This MCP server allows AI agents to interact with Atlassian products through a standardized interface. It provides extensive tools for:

• Confluence: Full CRUD operations, content search, space management, page management, comments, labels, and more using v2 REST API
• Jira: Issue management, project operations, transitions, comments, and comprehensive workflow support

Key Features

• Modern API Support: Uses Confluence v2 REST API with fallback to v1 for search functionality
• Comprehensive Toolset: 18+ Confluence tools and 8+ Jira tools covering all major operations
• Security & Privacy: Built-in PII filtering and SSL verification controls
• Flexible Configuration: Support for separate service URLs, authentication methods, and tool filtering
• Pagination Support: Cursor-based pagination for v2 APIs and offset/limit for v1 APIs
• Error Handling: Comprehensive error handling with detailed error messages
• Type Safety: Full TypeScript implementation with proper type definitions

Prerequisites

• Node.js (v16 or higher)
• npm or yarn
• Atlassian account with API token
• Docker (optional, for containerized deployment)

Installation

Standard Installation

1. Clone the repository:

   git clone https://github.com/yourusername/mcp-atlassian.git
   cd mcp-atlassian
   

2. Install dependencies:

   npm install
   # or using make
   make install
   

3. Create a .env file in the root directory with your Atlassian credentials:

   ATLASSIAN_HOST=https://your-domain.atlassian.net
   [email protected]
   ATLASSIAN_API_TOKEN=your-api-token
   

Docker Installation

1. Clone the repository:

   git clone https://github.com/yourusername/mcp-atlassian.git
   cd mcp-atlassian
   

2. Create a .env file as described above.

3. Build and run the Docker container:

   # Build the Docker image
   make docker-build
   
   # Run the Docker container
   make docker-run
   
   # Or use Docker Compose
   make docker-compose
   

Usage

Starting the Server

# Using npm
npm start

# Using make
make start

# Using Docker
make docker-run

This will start the MCP server, which will listen and respond on streamable HTTP.

MCP configuration

    "mcp-searchapi": {
      "name": "mcp-searchapi",
      "type": "streamable-http",
      "streamable": true,
      "url": "http://localhost:3007/mcp"
    }

Available Tools

Confluence Tools

Search & Discovery:

• search_confluence: Search for content in Confluence using v1 API with CQL

  • Parameters: searchText (string), spaceKey (string, optional), limit (number), start (number)

• search_confluence_pages_by_title: Search pages by title using v2 API