This project provides a custom web search agent built using the LangGraph framework, designed for users who need to integrate LLM-powered search capabilities into their applications. It offers a flexible and modular approach to building complex agentic workflows, simplifying the process of creating intelligent search tools.

How It Works

The agent leverages LangGraph to define and execute a stateful, multi-step workflow. This graph-based approach allows for explicit control over agent execution, enabling features like conditional routing, parallel processing, and state management between different tools and LLM calls. The core advantage is the ability to create robust and predictable agentic behavior by visualizing and managing the flow of information and actions.

Quick Start & Requirements

• Install: Clone the repository and install dependencies using pip install -r requirements.txt.
• Prerequisites: Python 3.11, Anaconda (recommended for environment management), API keys for Serper, OpenAI, Gemini, Claude, or Groq. Ollama can be used for local LLM inference.
• Setup: Requires configuring API keys in config.yaml. Running the frontend involves executing run_windows.ps1 or run_linux.sh.
• Docs: Agent Schema

Highlighted Details

• Supports multiple LLM providers (OpenAI, Gemini, Claude, Groq) and search APIs (Serper).
• Integrates with Ollama for local LLM model execution.
• Provides a frontend interface for interaction.

Maintenance & Community

No specific information on contributors, sponsorships, or community channels is provided in the README.

Licensing & Compatibility

The repository does not explicitly state a license.

Limitations & Caveats

The project requires obtaining and configuring API keys for various services, which may incur costs. The absence of a specified license could pose compatibility issues for commercial or closed-source use.