# LangGraph Research Brief Agent

## Overview
This repository implements a **LangGraph** agent that generates a short research brief for any given topic. The agent follows the assignment specification:

1. Create an outline of 4‑5 bullet points.
2. For each point perform one web search via Tavily and collect a concise note.
3. Synthesize all notes into a coherent brief (≈½–1 page).

The implementation uses:
- `langgraph` – graph orchestration
- `langchain-openai` – LLM calls to the BroJS GPT‑OSS model
- `langchain-tavily` – web search via Tavily API
- `python-dotenv` – environment variable loading (TAVILY_API_KEY, JOURNAL_MCP_PAT)

## Installation
```bash
pip install -r requirements.txt
```

## Usage
```bash
# Run the agent for a specific topic
python main.py "How to integrate LangGraph with Tavily"
```
The script prints:
- The generated outline
- Notes collected from web searches
- Final research brief

You can also import `run_brief` in your own code.

## Project structure
```
├── main.py          # Entry point and LangGraph implementation
├── requirements.txt # Dependencies
└── README.md        # Documentation
```

## Example outputs
1. **Topic:** *How to integrate LangGraph with Tavily*
2. **Topic:** *Best practices for building research briefs in AI*
3. **Topic:** *Using LangChain and LangGraph for educational projects*

Feel free to experiment with different topics.
""