get-gamesLABSCO SUMMARY
Shipping the endpoint schemas alongside the tools is what makes the natural-language questions actually work.
College football data is field-heavy, and a model guessing at column names produces plausible nonsense. Here the schema for each endpoint is available as a resource, so a question like which drives ended in a turnover inside the twenty gets translated against real field definitions. The prompt templates cover the analyses people ask for repeatedly — one game, one team, a season trend, a head-to-head, a rivalry — and the honest constraint is the rate limit rather than the data.
What it is
A Python server over the College Football Data API V2 that exposes the game, team, play and ranking endpoints, with schema documentation attached so a model knows what each field means.
What you get
- get-games and get-games-teams — results, scores and per-team game statistics
- get-records — team season records
- get-plays and get-play-stats — play-by-play data and individual play statistics
- get-drives — drive summaries and how they ended
- get-rankings — team rankings across the polls
- get-pregame-win-probability and get-advanced-box-score — win probability and advanced box score analytics
- Schema documentation for every endpoint as resources, so the model reads the field definitions before querying
- Prompt templates for common analyses: analyze-game, analyze-team, analyze-trends, compare-teams and analyze-rivalry
Requirements
A College Football Data API key in CFB_API_KEY — free to obtain. Python 3.11 or higher with uv. The free tier is rate limited; Patreon subscribers get higher limits, so wide season-long queries need to be spaced out.
Setup effort
Build from source — git clone https://github.com/lenwood/cfbd-mcp-server
USE CASES
01
Settling an argument about a specific game
Pull the box score and the drive summaries and look at what actually happened rather than what the final score suggests.
02
Finding the biggest upset of a season
Combine results with pregame win probability to rank surprises rather than margins.
03
Comparing two teams before they meet
Records, per-game team statistics and rankings for both, brought together in one answer.
04
Looking for a pattern in play calling
Query play-by-play data with the play statistics attached, filtered to the situations you care about.
9 tools
Listed in project docs. Your client may show a different list.
get-recordsget-games-teamsget-playsget-drivesget-play-statsget-rankingsget-pregame-win-probabilityget-advanced-box-scoreBEFORE YOU INSTALL
What to know first
Tools you get
9 tools named in the project's documentation.
Requires
- Python 3.11 or later
- CFB_API_KEY in your environment
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
You clone this one yourself.
Server setup
TERMINAL
git clone https://github.com/lenwood/cfbd-mcp-server
cd cfbd-mcp-server
uv venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate
uv pip install -e .Create a .env file in the project root containing CFB_API_KEY=<YOUR_CFB_API_KEY>
TERMINAL
uv run cfbd-mcp-serverMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"cfbd-mcp-server": {
"command": "uv",
"args": [
"--directory",
"<YOUR_PATH_TO_CFBD_MCP_SERVER>",
"run",
"cfbd-mcp-server"
],
"env": {
"CFB_API_KEY": "<YOUR_CFB_API_KEY>"
}
}
}
}<YOUR_CFB_API_KEY> is your College Football Data API key (the same key you put in .env as CFB_API_KEY; get one free at https://collegefootballdata.com/key), and <YOUR_PATH_TO_CFBD_MCP_SERVER> is the absolute path of your clone. The README's version of this block also sets PATH to a single python binary; it is left out here because PATH must be a list of directories, and a single file path there stops uv and python from being found.
