home / mcp / atlassian dc mcp server

Atlassian DC MCP Server

Provides MCP access to Jira, Confluence, and Bitbucket with per-service tokens and hosts.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "b1ff-atlassian-dc-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@atlassian-dc-mcp/jira"
      ],
      "env": {
        "JIRA_HOST": "your-jira-host",
        "BITBUCKET_HOST": "your-bitbucket-host",
        "JIRA_API_TOKEN": "your-token",
        "CONFLUENCE_HOST": "your-confluence-host",
        "BITBUCKET_API_TOKEN": "YOUR_TOKEN",
        "CONFLUENCE_API_TOKEN": "YOUR_TOKEN"
      }
    }
  }
}

The Atlassian DC MCP server lets your AI assistant interact with Jira, Confluence, and Bitbucket through a unified Model Context Protocol interface. It enables you to search, view, and create Jira issues; access and manage Confluence content; and work with Bitbucket repositories and code, all through a consistent, scriptable connection from your MCP client.

How to use

You configure MCP connections for each Atlassian Data Center product, then run the MCP client to start querying and acting on Jira, Confluence, and Bitbucket data. Your MCP client will communicate with the MCP server over the standard MCP endpoints, sending requests like issue lookups, content retrieval, and repository operations. Ensure each service has a valid API token and host information, and keep tokens secure.

How to install

Prerequisites you need before installing:

Node.js 18 or higher
npm 7 or higher (for workspaces support)
Atlassian Data Center or Cloud instance
API tokens for Jira, Confluence, and Bitbucket if you plan to connect to all three

Additional sections

Configuration details are provided in dedicated configuration blocks below. Use the examples that match your environment and populate tokens and host names accordingly. If you only use one Atlassian product, you can omit the others from your config.

Configuration and troubleshooting

The server supports two configuration approaches. You can use the standard local MCP configuration where you run the MCP connectors with npx and environment variables for each service. The alternative approach uses API base paths directly for each service. Choose the method that matches your security and deployment preferences.

Security and tokens

Store all API tokens securely. Do not expose tokens in logs or shared configurations. Use tokens only with the intended Atlassian service and restrict permissions to the minimum necessary for your use case.

Examples and notes

The following configuration blocks show how to enable Jira, Confluence, and Bitbucket MCP connections. Each block uses npx to launch the corresponding MCP package and passes authentication details via environment variables.

Available tools

jira

Search, view, and create Jira issues via MCP client integrations.

confluence

Access and manage Confluence content through MCP endpoints.

bitbucket

Interact with Bitbucket repositories and code via MCP connections.