Integrate Jira and Confluence with AI

MCP Atlassian is a Model Context Protocol server that connects AI assistants directly to Jira and Confluence, covering both Atlassian Cloud and Server/Data Center deployments. Once configured, an assistant can search issues with JQL, read and create Jira tickets, transition issue status, and search, read, create, and update Confluence pages and comments - all through natural-language requests instead of manual navigation between apps. Typical requests look like "find issues assigned to me in PROJ project," "search Confluence for onboarding docs," or "update the status of PROJ-123 to Done."

Use it when your workflow already lives in Jira and Confluence and you want an AI assistant to triage tickets, draft bug reports, pull sprint status, or search internal documentation without switching tabs. It is not a general-purpose ticketing replacement and adds no project-management features beyond what Jira and Confluence already expose through their APIs; teams on other trackers get no benefit from it, and Server/Data Center users should confirm their version meets the stated minimums (Jira 8.14+, Confluence 6.0+) before relying on it.

The server exposes 98 tools in total, split across the two products. Jira tools include jira_search (JQL-based search), jira_get_issue, jira_create_issue, jira_update_issue, and jira_transition_issue for moving tickets through workflow states. Confluence tools include confluence_search (CQL-based search), confluence_get_page, confluence_create_page, confluence_update_page, and confluence_add_comment. Authentication supports API tokens, Personal Access Tokens for Server/Data Center, and OAuth 2.0. It can run over HTTP transport with SSE or streamable-http, including multi-user setups, in addition to the default local mode.

Add it to your MCP client configuration (Claude Desktop, Cursor) with uvx:

{
  "mcpServers": {
    "mcp-atlassian": {
      "command": "uvx",
      "args": ["mcp-atlassian"],
      "env": {
        "JIRA_URL": "https://your-company.atlassian.net",
        "JIRA_USERNAME": "your.email@company.com",
        "JIRA_API_TOKEN": "your_api_token",
        "CONFLUENCE_URL": "https://your-company.atlassian.net/wiki",
        "CONFLUENCE_USERNAME": "your.email@company.com",
        "CONFLUENCE_API_TOKEN": "your_api_token"
      }
    }
  }
}

Server/Data Center deployments use JIRA_PERSONAL_TOKEN in place of the username/API-token pair. Docker, pip, and from-source installation methods are also documented, along with OAuth 2.0 as an alternative to API tokens. The project is MIT licensed and is not an official Atlassian product.

Teams and individual contributors already running Jira and/or Confluence - Cloud or self-hosted - who want an AI assistant to search tickets and docs, file and update issues, and keep Confluence pages current without leaving their chat or IDE. It is equally useful for support and dev teams triaging incoming bugs and for anyone maintaining internal documentation who wants faster search and edits.