EvoAgentX
ActiveDescription
Building a self-evolving ecosystem of AI agents with automatic optimization, role evolution, and multi-agent collaboration from single agent to complex systems.
Key Features
- Automatic multi-agent workflow construction from a single prompt description using structured task decomposition
- Built-in evaluation system with automatic scorers for task-specific agent behavior assessment
- Self-evolution engine that iteratively improves agent workflows through feedback loops and optimization
- Plug-and-play model compatibility with OpenAI, Claude, DeepSeek, Kimi, and local models via LiteLLM
- Comprehensive built-in tools for search, code execution, browser interaction, file I/O, and API calls
- Dual memory system supporting both short-term ephemeral and long-term persistent memory modules
Use Cases
💡 Automated optimization of agentic workflows using self-evolving algorithms and dataset-driven feedback
💡 Multi-agent system design for complex task decomposition and collaborative problem solving
💡 Human-in-the-loop AI workflows with interactive review and guidance checkpoints
💡 Rapid prototyping of LLM-based agent applications with automatic workflow generation
Categories
Quick Start
Install via pip: pip install evoagentx. Or install from source: pip install git+https://github.com/EvoAgentX/EvoAgentX.git. Set your API key as an environment variable. Import EvoAgentX and use the workflow builder to describe your goal — the framework automatically assembles a multi-agent workflow. Run the workflow and use built-in evaluators to measure performance.