A command-line fact-checking assistant that uses LLMs and web search to verify user claims. Built with LangGraph, and Tavily Search.
- Accepts user claims interactively via the terminal
- Searches the web for evidence using Tavily
- Uses a Google Gemini LLM to compare the claim against evidence
- Returns a verdict: TRUE, FALSE, or PARTIALLY TRUE, with detailed explanation and sources
- Clone the repository
- Install dependencies
pip install -r requirements.txt
- Set up API keys
- Create a
.envfile in the project root with the following:GOOGLE_API_KEY="<your-google-api-key>" TAVILY_API_KEY="<your-tavily-api-key>"
- Create a
Run the fact checker from the terminal:
python3 fact_checker.pyEnter a claim when prompted. The assistant will search for evidence and return a verdict with sources.
Type quit or exit to stop.
- Python 3.8+
- langchain
- langgraph
- langchain-tavily
- python-dotenv
MIT