This SDK enables the creation and execution of AI agents and collaborative agent swarms, supporting multiple LLM providers like OpenAI, Anthropic, and Gemini. It's designed for developers and researchers building complex AI workflows, offering a structured way to manage agent interactions and integrate external tools.

How It Works

The library utilizes a Python-based framework for defining agents with specific instructions and functions. Agents can be composed into swarms for collaborative tasks. It supports integrating external tools, such as web search or blockchain transaction lookups, and includes retrieval capabilities for processing documents via vector embeddings (supporting Chroma, Pinecone). Configuration is managed via TOML files, allowing customization of agent behavior and LLM provider settings.

Quick Start & Requirements

• Install via pip: pip install swarmzero
• Requires Python >= 3.11.
• An .env file with OPENAI_API_KEY (or other provider keys) is necessary.
• Detailed examples and API documentation are available at https://github.com/swarmzero/examples and http://localhost:8000/docs.

Highlighted Details

• Supports multiple LLM providers: OpenAI, Azure OpenAI, Anthropic, MistralAI, Gemini, Nebius, Ollama.
• Enables agent swarms for collaborative task execution.
• Integrates external tools and custom functions (e.g., Ethereum transaction lookup).
• Features document retrieval with support for basic, Chroma, Pinecone-serverless, and Pinecone-pod vector stores.
• Allows configuration of sample prompts for agent interaction.

Maintenance & Community

Project information and community links are not explicitly detailed in the README. The primary resource is the GitHub repository.

Licensing & Compatibility

The README does not specify a license. Compatibility for commercial use or closed-source linking is not detailed.

Limitations & Caveats

The README does not specify a license, which may impact commercial adoption. Some advanced features like specific LLM provider configurations or retrieval tool setups might require careful reading of example configurations.