kg-mcp

SnappStats/kg-mcp

3.2

If you are the rightful owner of kg-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.

The Model Context Protocol (MCP) server is a specialized server designed to facilitate communication and data exchange between machine learning models and various applications or platforms.

KG MCP

This project hosts an mcp that provides tools to extract/search and insert sports associated information into a knowledge graph, and to generate scout reports for players.

Local Development

Setup Python

It is recommended to use pyenv for managing versions of python.

  • Install pyenv:
brew install pyenv
  • Download and install the python version specified in 
pyenv install
  • Start using the installed python version:
pyenv local

Environment Variables

We use Doppler for most environment variables.

  • Log in to Doppler in your web browser. The credentials for Doppler are in the 1Password vault.

  • Install the Doppler CLI:

brew install dopplerhq/cli/doppler
# or make install-doppler
  • Log in to the Doppler CLI:
doppler login
  • Configure environment variables for an application:
doppler setup

In general, keep environment variables -- especially secrets -- in Doppler. Use .env only for local overrides.

Run the mcp server

uv run poe dev

Test locally using an mcp client

  1. Make sure your .env file has the following variables set:
KG_MCP_SERVER=http://127.0.0.1:8001/mcp
  1. Run the MCP server locally (see above).
  2. In another terminal, run the curation bot, which behaves like a root agent:
uv run python curation_bot.py

Test Coverage

Run all tests

uv run poe test

Execute a specific test

uv run poe test_file tests/<path_to_file>.py

All tests should go inside the tests folder sitting at the root of the directory

Deploy

  1. Merge to main branch, and git push.
  2. Check that it shows up at: https://kg-mcp-762632998010.us-central1.run.app