AI Agent Framework
3.4k 2026-04-18
aiming-lab/MetaClaw
An AI agent framework enabling continuous meta-learning and evolution from live conversations, requiring no GPU.
Core Features
Continuous Meta-Learning from Live Interactions
No GPU Cluster Required for Operation
Automatic Skill Evolution and Improvement
Persistent Cross-Session Memory and Context
One-Click Deployment and Easy Setup
Quick Start
metaclaw setup && metaclaw startDetailed Introduction
MetaClaw is an innovative AI agent framework designed to empower agents with the ability to meta-learn and evolve directly from real-world conversations. Inspired by biological learning, it transforms every interaction into a valuable learning signal, allowing agents to continuously improve their skills in deployment rather than relying solely on offline training. A key advantage is its efficiency, operating without the need for GPU clusters, making advanced AI agent development accessible and scalable. It features persistent memory, multi-agent support, and simplified one-click deployment.