jira-promps-mcp-server

jira-promps-mcp-server

0

The Jira Prompts MCP Server is a Model Context Protocol server designed to generate prompts from Jira content. It uses the `pycontribs/jira` API and is ideal for tools supporting only MCP prompts, aiding in contexts like Zed's AI assistant.

Jira Prompts MCP Server

This repository provides a Model Context Protocol (MCP) server that offers several commands for generating prompts or contexts from Jira content.

This repository draws significant inspiration from the MarkItDown MCP server and the example Git MCP server. It also incorporates design and code elements from the mcp-atlassian repository. The main differences between this repository and mcp-atlassian are that it uses pycontribs/jira instead of atlassian-api/atlassian-python-api to interact with the Jira API, and it focuses on providing prompts rather than tools. The latter makes it especially useful when working with tools that support only MCP prompts, such as Zed's AI assistant.

Here's another MCP server project of mine: ceshine/git-prompts-mcp-server

Changelog

0.1.0

  • Migrate from the low-level mcp package to the FastMCP package.
  • Add a CLI for testing the server.

0.0.1

The initial release with two prompts implemented: jira-issue-brief and jira-issue-full.

Installation

Manual Installation

  1. Clone this repository
  2. Install dependencies: uv sync --frozen

Usage

As a MCP Server for Zed Editor

Add the following to your settings.json:

"context_servers": {
  "git_prompt_mcp": {
    "command": {
      "path": "uv",
      "args": [
        "--directory",
        "/path/to/local/jira_prompts_mcp_server",
        "run",
        "jira-prompts-mcp-server",
        "https://my-company.atlassian.net", // Jira base URL
        "your_jira_account@example.com",  // Jira username
        "your_api_key"  // Jira API token (https://id.atlassian.com/manage-profile/security/api-tokens)
      ]
    },
    "settings": {}
  }
}
Commands

The server responds to the following commands:

  1. jira-issue-brief <issue-key>: Retrieves the core fields of a Jira issue. Requires the issue key (e.g., PROJ-123) as an argument.
  2. jira-issue-full <issue-key>: Retrieves the core fields, comments, linked issues, and subtasks of a Jira issue. Requires the issue key as an argument.

Examples:

  1. /jira-issue-brief PROJ-123
  2. /jira-issue-brief PROJ-155

Testing the server using the CLI

Prerequisities: configuring the required environment variables (JIRA_URL, JIRA_USERNAME, JIRA_API_TOKEN)

You can test the

  • uv run python -m jira_prompts_mcp_server.cli jira-brief BOOM-1234
  • uv run python -m jira_prompts_mcp_server.cli jira-full BOOM-1234

License

MIT License. See for details.