Overview
What is Shortcut.com MCP Server?
An implementation of the Model Context Protocol (MCP) server that allows Claude and other MCP-compatible AI assistants to interact with Shortcut.com’s ticket management system. It enables listing, searching, creating, updating stories, and adding comments, as well as retrieving workflow states and projects.
How to use Shortcut.com MCP Server?
Install Python 3.10+, clone the repository, create a virtual environment, install dependencies, and set up a .env file with your Shortcut API token. Run the server with python -m src.server. To use with Claude Desktop, add the server configuration to claude_desktop_config.json with the SHORTCUT_API_TOKEN environment variable.
Key features of Shortcut.com MCP Server
- List and search stories (tickets) in Shortcut.
- Get detailed information about specific stories.
- Create and update stories.
- Add comments to stories.
- Retrieve workflow states and projects.
- Provides pre-built bug report and feature request templates.
Use cases of Shortcut.com MCP Server
- Automate bug report creation from AI conversations.
- Quickly search and retrieve story details without leaving the chat.
- Update story statuses or add comments via natural language commands.
- Generate structured feature request templates for new work items.
FAQ from Shortcut.com MCP Server
What are the prerequisites for Shortcut.com MCP Server?
Python 3.10+ and a Shortcut.com API token are required.
How do I configure Shortcut.com MCP Server with Claude Desktop?
Edit the Claude Desktop configuration file (claude_desktop_config.json) and add the server entry with the command, args, and env containing the SHORTCUT_API_TOKEN. Then restart Claude Desktop.
What MCP capabilities does the server provide?
It exposes resources (shortcut://stories, shortcut://story/{story_id}), tools (list, search, create, update, add comment, list workflow states, list projects), and prompts (create bug report, create feature request).
How can I add new capabilities to the server?
Add API methods to shortcut_client.py, define Pydantic models in utils.py if needed, and implement MCP functionality using decorators (@mcp.resource, @mcp.tool, @mcp.prompt) in server.py.
Is the server open source and how is it licensed?
Yes, contributions are welcome. The project is licensed under the MIT License.
