Atlassian MCP Server
An MCP server hosted by Atlassian at `mcp.atlassian.com`, providing MCP-based access to Atlassian products and data streams over SSE.
About this tool
Atlassian MCP Server
Remote MCP Server hosted by Atlassian that connects Jira and Confluence Cloud to MCP-compatible clients (LLMs, IDEs, and agent platforms) over SSE.
Features
- Remote MCP bridge between Atlassian Cloud (Jira, Confluence) and external MCP clients.
- Real-time interaction with Jira and Confluence data via Server-Sent Events (SSE) endpoint:
- Endpoint:
https://mcp.atlassian.com/v1/sse.
- Endpoint:
- OAuth 2.1 / 2.0–based authorization for secure, scoped access.
- Respects Atlassian permissions:
- Only data the user can see in Jira/Confluence is accessible.
- Actions comply with project/space-level roles.
- Summarize and search Jira and Confluence content from compatible tools without switching apps.
- Create and update Jira issues and Confluence pages using natural language commands.
- Bulk task processing, such as:
- Generating multiple tickets from meeting notes.
- Creating tasks from product specs or other documents.
- Works with multiple client types:
- LLM-based tools (e.g., Claude for Teams).
- MCP-compatible IDE integrations (e.g., Claude desktop, VS Code, Cursor).
- Custom MCP-compatible clients via the defined SSE endpoint.
- Supports local MCP proxy setups using a Node.js-based
mcp-remoteclient.
How It Works
- A supported MCP client connects to the Atlassian MCP Server SSE endpoint:
https://mcp.atlassian.com/v1/sse. - The server triggers a secure, browser-based OAuth authorization flow.
- After authorization, the client can:
- Stream contextual data to the server.
- Receive real-time responses from Jira and Confluence.
- All requests and actions are constrained by the user’s existing Atlassian Cloud permissions and OAuth token scopes.
Requirements
Cloud-based Setup
- An Atlassian Cloud site with Jira and/or Confluence.
- Access to an AI client that supports MCP (e.g., Claude for Teams).
- A modern web browser to complete the OAuth authorization flow.
Desktop / Local Client Setup
- An Atlassian Cloud site with Jira and/or Confluence.
- A supported IDE or client, such as:
- Claude desktop
- VS Code
- Cursor
- Any custom MCP-compatible client
- Node.js v18+ installed to run the local MCP proxy (
mcp-remote). - A modern browser to complete OAuth login.
Data & Security
- All traffic is encrypted via HTTPS using TLS 1.2 or later.
- OAuth 2.0 / 2.1 is used for secure authentication and access control.
- Access to Jira and Confluence data is governed by the user’s existing permissions (project/space-level roles and scopes).
- OAuth tokens are scoped and session-based.
Rate Limits
- Standard plan:
- Moderate usage thresholds (no specific numeric limit provided).
- Premium / Enterprise plans:
- Higher usage quotas, including around 1,000 requests per hour plus additional per-user limits.
Pricing
The documentation describes Standard and Premium/Enterprise usage tiers in terms of rate limits but does not specify monetary pricing or subscription costs. Plans and pricing details would depend on the underlying Atlassian plan associated with your site.
Loading more......
Information
Categories
Similar Products
6 result(s)A curated GitHub repository that maintains a concise list of MCP (Model Context Protocol) servers. It serves as a directory of available MCP servers, aiming to help developers quickly discover and explore MCP-compatible server implementations, tools, and related resources.
A curated GitHub repository listing remote Model Context Protocol (MCP) servers, including categories, authentication details, and usage examples for integrating MCP servers with LLM clients. It serves as a central directory for discovering and contributing OAuth 2.0–enabled remote MCP servers.
A GitHub-hosted, automatically ranked and weekly updated curated list of approximately 400 Model Context Protocol (MCP) servers, grouped into 34 categories and scored using project-quality metrics collected from GitHub and package managers. It serves as a centralized directory for discovering top MCP servers and their ecosystems.
A web directory of MCP clients and servers that helps developers discover, compare, and choose the right Model Context Protocol tools.
An open-source collection of example MCP servers (TypeScript and Python) managed by Anthropic and the Model Context Protocol community, including servers for memory, filesystem access, Git, GitHub, Postgres, and more.
Atlassian’s MCP-based integration for IDEs, used by Rovo to connect tools like Jira, Confluence, and other Atlassian products to AI assistants. The Atlassian CLI MCP server can be registered in environments like Warp so models can query and act on Atlassian project data from within the editor.