# atlassian-mcp-server [Verified] [Health: Active]

**Category:** 🔄 Version Control  
**Repository:** https://github.com/atlassian/atlassian-mcp-server  
**GitHub Stars:** 1151  
**Views:** 11  
**Installs:** 0  
**Upvotes:** 1  
**Directory Page:** https://allmcps.com/mcp/atlassian-mcp-server

## Description
Official remote MCP server for Atlassian. Securely connect Jira, Confluence, Jira Service Management, Bitbucket, and Compass to Claude, ChatGPT, Cursor, VS Code, and other AI tools using OAuth 2.1 or API tokens.

## Claude Desktop Quick Installation
Remote MCP endpoint (confidence: high). Install path detected from listing signals. Add as a URL/SSE server in your client:

```json
"mcpServers": {
  "atlassian-mcp-server": {
    "url": "https://mcp.atlassian.com/v2/mcp"
  }
}
```

## Documentation

## What atlassian-mcp-server does

atlassian-mcp-server is Atlassian’s hosted MCP service for connecting AI clients with Atlassian Cloud data and actions. It covers Jira, Confluence, Jira Service Management, Bitbucket, Compass, and Loom, along with platform data such as Projects, Goals, Teams, Focus, Talent, and the Teamwork Graph.

The server supports searching and summarizing Jira, Jira Service Management, Confluence, Bitbucket, Projects, Goals, and related data. It can also retrieve and review Loom recordings, create or update work items and pages, and turn material such as meeting notes or specifications into work items.

## How it works

An MCP-compatible AI client connects to Atlassian’s remote service rather than running a local server process. The recommended endpoint for new setups is:

`https://mcp.atlassian.com/v2/mcp`

After the connection is configured, the AI client selects the relevant Atlassian actions from natural-language requests. Authentication uses OAuth 2.1 or an API token, and operations follow the connected user’s existing Atlassian permissions.

The service is hosted by Atlassian Cloud. Existing v1 connections begin exposing v2 tools automatically, although some clients may need cached client IDs or `.well-known` credentials cleared if authentication becomes incompatible after the transition.

## Setup and configuration

Add the remote MCP endpoint to a compatible client using its native MCP installation flow or configuration interface. The repository provides one-click installation links for Cursor and VS Code, using the v2 URL.

Authentication can be configured with OAuth 2.1 or an API token. The selected method determines how the client accesses Atlassian resources and ensures actions remain subject to the user’s permissions. New installations should target the v2 endpoint because it exposes additional tools and products.

## Tools and capabilities

- Search and summarize Jira issues, Jira Service Management data, Confluence content, Bitbucket data, Projects, Goals, and other Atlassian information.
- Retrieve and review Loom video and meeting recordings.
- Create and update Jira work items.
- Create and update Confluence pages.
- Generate work items from meeting notes or specifications.
- Access Compass and broader Atlassian platform data through the Teamwork Graph.

## Limitations and notes

This service is designed for Atlassian Cloud and requires an MCP-compatible external AI client. Access is constrained by the connected user’s Atlassian permissions, so the AI client cannot use resources that user cannot access.

The recommended v2 service may expose different tools from older v1 connections. Clients with incompatible cached authentication metadata may require local credential-cache cleanup before they can authenticate again.

The repository is licensed under Apache 2.0.

_Full upstream README: https://allmcps.com/mcp/atlassian-mcp-server/readme_