Project TTV

This repository contains instructions for setting up and running an MCP service that can be published to PyPI.

Setup Instructions

1. Initialize UV Environment

uv init --build-backend "hatchling"
uv add "mcp[cli]"

2. Create Project Structure

touch src/project_ttv_mcp/project_ttv_mcp.py

Package Development

To turn this MCP service into a Python package that can be published to PyPI:

1. Update __init__.py to include a main() function that runs MCP
2. Create a __main__.py script that calls the main function and serves as an entry point

Running the Service

Local Development

You can run this MCP service as a module using either of these methods:

uv run -m project-ttv-mcp

Configuration Options

Local Configuration

{
  "mcpServers": {
    "Demo": {
      "command": "uv",
      "args": [
        "--directory",
        "/Users/ram/Documents/J360-Repo/Project-TTV-MCP",
        "run",
        "-m",
        "project_ttv_mcp"
      ]
    }
  }
}

GitHub Repository Configuration

Using uvx to install dependencies in a temporary environment:

{
  "mcpServers": {
    "Demo": {
      "command": "uvx",
      "args": [
        "git+https://github.com/TrelisResearch/ADVANCED-inference#subdirectory=trelis-mcp/pypi-mcp-test"
      ]
    }
  }
}