Model Context Protocol (MCP) Atlassian server provides a unified integration layer for Atlassian Confluence and Jira across Cloud, Server, and Data Center deployments. Delivered as a Docker image, it enables AI assistants and IDEs to perform read and write operations on Jira issues and Confluence content, including searching, retrieving pages, creating/updating pages, and managing issues. It exposes Jira and Confluence tools such as jira_get_issue, jira_search, confluence_search, confluence_get_page, and confluence_create_page, among others. The MCP supports multiple authentication modes: API tokens for Cloud, Personal Access Tokens for Server/DC, and OAuth 2.0 Cloud with a setup wizard; there is also a Bring Your Own Token (BYOT) option that bypasses internal token management. It includes multi-cloud OAuth support for multi-tenant use, allowing per-request cloud context via headers. Additional features cover proxies and per-service overrides, custom HTTP headers, and two transport modes (SSE at /sse or streamable-http at /mcp). Configuration can be done via environment variables or an env-file, and the server emphasizes tool filtering, read/write controls, and IDE integrations for seamless development workflows. Tools are listed to cover common Jira/Confluence operations.