Fantasy Football AI
¶
Using AI to dominate your fantasy football league. It's like having a secret weapon, but for nerds.
This project leverages Google's Gemini AI to provide data-driven insights and analysis for your fantasy football league. The behavior and instructions for the AI agents are defined in AGENTS.md
files throughout the repository, with the main guidelines available at AGENTS.md. It helps you make informed decisions on player stats, team analysis, and trade suggestions.
Development Version
This project is currently in a development stage. Features and configurations are subject to change, and breaking changes may be introduced at any time.
TL;DR¶
-
Initial Setup
-
Weekly Update
-
Draft Preparation
Features¶
- Player Stat Analysis: Downloads and processes weekly player statistics from the
nfl_data_py
library, including dynamic ADP data from a reliable API. - Customizable Scoring Rules: Define your league's scoring rules in
config.yaml
for accurate fantasy point calculations. - Team Identification: Easily identify and save your team from a list of all teams in the league.
- Report Generation: Generates comprehensive Markdown reports summarizing draft recommendations, bye week conflicts, and trade suggestions.
- Team Analysis: Uses the Gemini AI to analyze your team's roster and provide feedback.
- Trade Suggestions: Identifies "buy-low" and "sell-high" trade candidates based on weekly performance versus season averages.
- Waiver Wire Gems: Identifies high-upside waiver wire candidates based on recent usage trends and underperforming fantasy points.
- Live Draft Assistant: Provides an interactive tool for real-time draft suggestions based on VBD and team needs.
- Available Player Downloads: Fetches a list of all available players from your ESPN league.
- League Settings: Downloads your current league settings, including scoring and roster rules.
- Customizable Tasks: Uses
Taskfile
to easily run common commands for linting, installing dependencies, downloading data, and running analysis.
License¶
Author¶
This project was started in 2025 by Nicholas Wilde