Overview

The Atlassian Rovo MCP Server is the official remote Model Context Protocol (MCP) server from Atlassian. It acts as a secure, cloud-hosted bridge between your Atlassian Cloud site (Jira, Confluence, and Compass) and external AI tools, IDEs, and agent platforms. No local installation is required — it uses a simple OAuth flow and operates as a proxy that respects your existing permissions.

Launched in 2025 and hosted on Cloudflare, it brings Atlassian context directly into tools like Anthropic Claude, Cursor, VS Code, and other MCP-compatible clients.

Key Features

Remote & Zero-Setup: Fully hosted remote MCP server — connect via https://mcp.atlassian.com/v1/mcp (SSE deprecated after June 2026).
Secure Proxy: Does not store or cache your Jira/Confluence data; acts only within the signed-in user's permissions.
Read & Write Operations: Search, summarize, create, update, and bulk-manage Jira issues, Confluence pages, and Compass components.
Natural Language Support: AI agents can perform complex tasks like "Find all open bugs in Project X" or "Create a story and link it to a Confluence page".
Broad Client Compatibility: Works with Claude Desktop, Cursor, VS Code (via built-in MCP support), and other MCP clients.
Enterprise Ready: Rate limits based on your Jira/Confluence plan; complies with Atlassian Privacy Policy and Data Processing Addendum.

Supported Capabilities

Jira: Search issues with JQL, get issue details, create/update issues, add comments, manage sprints, watchers, and more.
Confluence: Search pages, summarize content, create/update pages.
Compass: Access components and related data.
Cross-tool Workflows: Link issues to pages, summarize work across products, bulk operations.

Use Cases

AI-Powered Project Management: Let Claude or Cursor understand your sprint, create tickets, or update status from natural language prompts.
Context-Aware Coding: Bring Jira issue details and Confluence documentation directly into your IDE for better code suggestions and task automation.
Automation & Orchestration: Build multi-agent workflows that interact with Jira alongside GitHub, Azure DevOps, or other MCP servers.
Knowledge Retrieval: Quickly summarize Confluence pages or find related issues without leaving your AI chat.
Enterprise AI Adoption: Safely extend powerful LLMs into regulated environments while maintaining data control.

Getting Started

Ensure you have an Atlassian Cloud site with Jira, Confluence, or Compass.
In your MCP-compatible client (Claude, Cursor, VS Code, etc.), add the server endpoint: https://mcp.atlassian.com/v1/mcp.
Authenticate via the secure browser-based OAuth 2.1 flow.
Start using natural language commands — the AI will automatically use the available Jira/Confluence tools.

Full setup guides are available in Atlassian Support documentation.

Community & Self-Hosted Alternatives

For on-prem Jira Server/Data Center or additional customization, community projects exist such as mcp-atlassian and various open-source Jira-only MCP servers on GitHub.

Resources

Official Page: https://www.atlassian.com/platform/remote-mcp-server
Getting Started: https://support.atlassian.com/atlassian-rovo-mcp-server/docs/getting-started-with-the-atlassian-remote-mcp-server/
GitHub Repository: https://github.com/atlassian/atlassian-mcp-server
Use Guide: https://support.atlassian.com/atlassian-rovo-mcp-server/docs/use-atlassian-rovo-mcp-server/

The Atlassian Rovo MCP Server is the recommended way for most Jira Cloud users to integrate with the growing MCP ecosystem.

Atlassian Rovo MCP Server (Jira & Confluence)