Jira Weekly Reporter

You run a FastMCP server that connects to Jira to generate weekly reports on issue activity and optionally summarizes them using the client’s LLM. It exposes a tool named generate_jira_report that you can call from an MCP client to fetch Jira data, filter it with a JQL, and manage the size of the results.

How to use

You will call the generate_jira_report tool from an MCP client to fetch Jira issues based on a JQL query. The tool can optionally summarize the report using the client’s LLM by enabling summarization.

How to install

Prerequisites are Python 3.10 or later and a Jira instance you can access with an API token. You should have the FastMCP CLI installed and available in your system’s PATH.

# Install dependencies and the server package
uv pip install fastmcp "jira[cli]" python-dotenv httpx anyio

# Or with pip directly
pip install fastmcp "jira[cli]" python-dotenv httpx anyio

Create a .env file in the same directory as jira_reporter_server.py and add your Jira connection details.

# .env
JIRA_URL=https://your-domain.atlassian.net
JIRA_USERNAME=your_email@example.com
JIRA_API_TOKEN=your_api_token_or_pat

Run the server directly or via the FastMCP CLI.

# Run with Python directly
python jira_reporter_server.py

# Or run with FastMCP
fastmcp run jira_reporter_server.py

# Run with SSE for remote access
fastmcp run jira_reporter_server.py --transport sse --port 8001

Additional sections

Configuration and security notes: store Jira credentials in a local *.env* file and ensure it is not committed to version control. For Jira Cloud, generate an API token; for Jira Server/Data Center, generate a Personal Access Token. The server reads connection details from the environment, so keep the *.env* file secure and exclude it from repositories.

MCP tool details: the server exposes a tool named generate_jira_report. It accepts a JQL query, an optional project key to scope results, a maximum number of issues to include, and an option to summarize the report using the client’s LLM.

Troubleshooting tips: ensure the Jira API token has the necessary permissions, verify the Jira URL is reachable from the host running the server, and confirm the environment variables are loaded correctly in the running process. If you encounter networking errors, check firewall rules and proxy settings. If the Jira library call blocks, the server will offload it to a separate thread via asyncio.to_thread to avoid blocking the event loop.

MCP configuration (server details)

{
  "mcpServers": [
    {
      "name": "jira_report",
      "type": "stdio",
      "command": "fastmcp",
      "args": ["run", "jira_reporter_server.py"],
      "env": [
        {"name": "JIRA_URL", "value": "https://your-domain.atlassian.net"},
        {"name": "JIRA_USERNAME", "value": "your_email@example.com"},
        {"name": "JIRA_API_TOKEN", "value": "your_api_token_or_pat"}
      ]
    }
  ]
}

Environment variables are required to connect to Jira. You can provide them in a cluster-wide environment or via the server’s own environment map shown above.

Available tools

generate_jira_report

Generates a Jira issue report based on a JQL query, with optional summarization via the client’s LLM. Parameters include the JQL, project scoping, maximum results, and whether to summarize.