MCP Atlassian Server

You use MCP Atlassian to connect AI-powered automation with Atlassian Confluence and Jira, across Cloud and Server deployments. It lets you search, create, update, and manage Jira issues and Confluence pages through a programmable MCP server that your IDE or tooling can talk to.

How to use

You run a MCP Atlassian server and connect your IDE or client to it, either through a standard HTTP transport or a local stdio-based setup. You can operate in single-user mode with fixed credentials or enable multi-user authentication so each user provides their own Atlassian access. Once connected, you can perform actions such as searching Jira issues, creating or updating Confluence pages, and updating Jira issues from your assistant or automation scripts.

How to install

Prerequisites: you need Docker installed on your machine to run the MCP Atlassian server as a container.

Install and start the server using the pre-built image. Pull the image and run it with your preferred configuration.

docker pull ghcr.io/sooperset/mcp-atlassian:latest

Configuration and usage notes

The server supports multiple configuration methods. You can pass environment variables directly, supply an env-file, or use explicit JSON configuration blocks to run the container. The following configurations illustrate common patterns you will deploy in real environments.

{
  "mcpServers": {
    "mcp-atlassian": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "-p", "8080:8080",
        "-v", "${HOME}/.mcp-atlassian:/home/app/.mcp-atlassian",
        "ghcr.io/sooperset/mcp-atlassian:latest",
        "--oauth-setup",
        "-v"
      ],
      "env": {
        "JIRA_URL": "https://your-company.atlassian.net",
        "CONFLUENCE_URL": "https://your-company.atlassian.net/wiki"
      }
    }
  }
}

OAuth 2.0 standard flow (Cloud)

To enable the built-in OAuth client for Atlassian Cloud, you complete the setup wizard and persist tokens locally. This is the standard flow for most users.

Run the setup wizard to configure OAuth and start token persistence.

docker run --rm -i \
   -p 8080:8080 \
   -v "${HOME}/.mcp-atlassian:/home/app/.mcp-atlassian" \
   ghcr.io/sooperset/mcp-atlassian:latest --oauth-setup -v

BYOT: Bring Your Own OAuth Access Token (Cloud)

If you manage OAuth tokens outside this server, you can provide an existing access token directly. Tokens are not refreshed by the MCP server in this mode.

{
  "mcpServers": {
    "mcp-atlassian": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "-e", "JIRA_URL",
        "-e", "CONFLUENCE_URL",
        "-e", "ATLASSIAN_OAUTH_CLOUD_ID",
        "-e", "ATLASSIAN_OAUTH_ACCESS_TOKEN",
        "ghcr.io/sooperset/mcp-atlassian:latest"
      ],
      "env": {
        "JIRA_URL": "https://your-company.atlassian.net",
        "CONFLUENCE_URL": "https://your-company.atlassian.net/wiki",
        "ATLASSIAN_OAUTH_CLOUD_ID": "YOUR_CLOUD_ID",
        "ATLASSIAN_OAUTH_ACCESS_TOKEN": "YOUR_PRE_EXISTING_OAUTH_ACCESS_TOKEN"
      }
    }
  }
}

Single service configurations

If you want to run a single service configuration for a specific Atlassian product, you can tailor the environment accordingly. The examples show how to run for Confluence Cloud, Confluence Server/Data Center, Jira Cloud, and Jira Server/Data Center.

{
  "mcpServers": {
    "mcp-atlassian": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "-e", "CONFLUENCE_URL",
        "-e", "CONFLUENCE_USERNAME",
        "-e", "CONFLUENCE_API_TOKEN",
        "ghcr.io/sooperset/mcp-atlassian:latest"
      ],
      "env": {
        "CONFLUENCE_URL": "https://your-company.atlassian.net/wiki",
        "CONFLUENCE_USERNAME": "your.email@example.com",
        "CONFLUENCE_API_TOKEN": "your_confluence_api_token"
      }
    }
  }
}

Proxy, headers and advanced configuration

You can route API requests through HTTP/HTTPS/SOCKS proxies and add custom HTTP headers to all requests. Use environment variables to set proxies and headers per service.

{
  "mcpServers": {
    "mcp-atlassian": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e", "HTTP_PROXY",
        "-e", "HTTPS_PROXY",
        "-e", "NO_PROXY",
        "ghcr.io/sooperset/mcp-atlassian:latest"
      ],
      "env": {
        "HTTP_PROXY": "http://proxy.internal:8080",
        "HTTPS_PROXY": "http://proxy.internal:8080",
        "NO_PROXY": "localhost,.your-company.com"
      }
    }
  }
}

Troubleshooting and debugging

If you encounter authentication or SSL issues, verify tokens and SSL verification flags. Enable verbose logging to inspect request headers and token usage.

Tools and capabilities

The MCP server provides a set of tools to read and write Jira and Confluence data. Jira tools include getting issues, searching, creating, updating, and commenting; Confluence tools cover page search, creation, and updates.

Security and best practices

Never share API tokens or tokens. Keep environment files secure and private. Use the server's access controls to limit who can initiate actions against your Atlassian data.

Available tools

jira_get_issue

Get details of a specific Jira issue by key

jira_search

Search Jira issues using JQL

jira_create_issue

Create a new Jira issue

confluence_search

Search Confluence content using CQL

confluence_get_page

Get content of a Confluence page by ID