lang-test

Multi-agent AI research system built with LangChain and LangGraph. Automatically generates search queries, spawns parallel search agents, filters results by relevance, and synthesizes comprehensive research reports.

Features

• 🧠 Intelligent Query Generation - Automatically creates 3-5 diverse search queries from any topic
• 🚀 Parallel Search Agents - Spawns multiple agents to search concurrently
• 🔬 AI-Powered Filtering - Analyzes and ranks results by relevance (0-10 scale)
• 📊 Comprehensive Synthesis - Combines findings into structured reports with sources
• 💾 MongoDB Storage - Saves research results for future reference

Prerequisites

• Node.js
• Ollama running locally
• SearxNG running locally
• MongoDB running locally

Setup

1. Install dependencies:

yarn install

2. Copy .env.example to .env and configure:

cp .env.example .env

3. Run:

yarn dev