atlassian

This Atlassian MCP Server provides a unified JSON-RPC interface to interact with Jira Server, Confluence Server, Bitbucket, and Bamboo through a single MCP server. The server supports two transport modes: stdio (default) for process-based integrations and HTTP for remote integrations. Clients communicate using MCP methods such as initialize, tools/list to discover available tools, and tools/call to execute specific tool operations (e.g., jira_get_issue, confluence_create_page, bamboo_trigger_build). The server translates MCP requests into Atlassian API calls and returns structured results or mapped errors. Typical usage involves starting the server with a configuration that enables the tools you need and then issuing JSON-RPC requests to call the desired operations. Each tool exposes a comprehensive set of operations, from Jira issue management to Bamboo build control, all accessible via the same MCP surface.

To leverage the HTTP transport, configure the server with transport.type set to http and expose a host/port. Clients can then POST JSON-RPC requests to the endpoint and receive responses. For stdio transport, you typically spawn the server as a child process and communicate via stdin/stdout, which is convenient for scripting and integration with other CLI tools. The server includes authentication configuration per tool (basic or token) and provides robust error mapping to MCP error codes for configuration, authentication, API, network, and rate-limiting scenarios.

Prerequisites:

• Go 1.24.1 or later (for building from source) or use the provided Docker setup
• Access credentials for Jira, Confluence, Bitbucket, and Bamboo as needed

Option A: Build from source (Go)

1. Clone the repository:

git clone <repository-url>
cd atlassian-mcp-server

2. Install dependencies:

go mod download

3. Build the server:

go build -o atlassian-mcp-server main.go

4. Create a config.yaml file in the working directory and run:

./atlassian-mcp-server

Option B: Docker (recommended for quick start)

1. Build the Docker image:

docker build -t atlassian-mcp-server .

2. Run with your configuration mounted:

docker run -v $(pwd)/config.yaml:/app/config.yaml atlassian-mcp-server

3. If using HTTP transport, expose the port (e.g., 8080):

docker run -p 8080:8080 -v $(pwd)/config.yaml:/app/config.yaml atlassian-mcp-server

Note on configuration:

• Copy the example config.yaml to config.yaml and edit to enable the Atlassian tools you need, set base_url, authentication, and transport type (stdio or http).
• For HTTP, specify host and port in the config as described in the README.

Tips and common issues:

• Ensure the base_url for each Atlassian tool is reachable from the server host and that credentials (basic or token) are correct.
• If using HTTP transport, ensure the host/port are not blocked by firewalls and that TLS considerations are handled if needed.
• The MCP error codes include authentication (-32002) and API (-32003) errors; enable verbose logging to diagnose misconfigurations quickly.
• When running in Docker, you may need to pass environment variables or mount volumes for config and credentials securely.
• The server supports a wide range of Jira/Confluence/Bitbucket/Bamboo operations; consult the Available Tools section of the README to discover exact operation names and required arguments.