- rea by morluto, 460k stars. Agent-driven reverse engineering from app behavior down to native binaries
- openclaw by openclaw, 394k stars. Personal AI assistant that runs tasks on your own machine across platforms
- deepseek-harness by deepseek-ai, 281k stars. Plugin-based agent harness where every feature is a plugin
- hermes-agent by NousResearch, 263k stars. Personal agent that builds memory and skills over time
- AutoGPT by Significant-Gravitas, 188k stars. Platform for building and running autonomous AI agents
- agency-agents by msitarzewski, 168k stars. Specialized agent personas with defined processes for many business roles
- dify by langgenius, 161k stars. Platform for agentic workflows and RAG pipelines with model and tool support
- langflow by langflow-ai, 157k stars. Visual builder for deploying AI agents and workflows built on React Flow
- langchain by langchain-ai, 150k stars. Platform for building LLM agents and applications in Python
- TradingAgents by TauricResearch, 123k stars. Framework where LLM agents collaborate on stock trading decisions
- claude-mem by thedotmack, 117k stars. Stores agent session activity, compresses it and reinjects context later
- paperclip by paperclipai, 114k stars. Open-source app for managing and coordinating agents for work
- Agent-Reach by Panniantong, 112k stars. CLI giving agents read and search access to Twitter, Reddit, YouTube and more
- MiroFish by 666ghj, 86k stars. Multi-agent swarm engine for predicting outcomes from a knowledge graph
- lobehub by lobehub, 86k stars. Workspace for building, scheduling and tracking a team of AI agents
- deer-flow by bytedance, 85k stars. Long-running agent harness with sandboxes, memory, tools and subagents
- ruflo by ruvnet, 82k stars. Multi-agent swarm runtime with coordination, memory and autonomous workflows
- ai-agents-for-beginners by microsoft, 80k stars. 18 lessons with notebooks on building AI agents from scratch
- MetaGPT by FoundationAgents, 71k stars. Multi-agent framework that assigns software company roles to LLM agents
- ai-hedge-fund by virattt, 65k stars. Python project with AI agents acting as an investment team
- hindsight by vectorize-io, 61k stars. Memory system for AI agents that learns from past interactions
- crewAI by crewAIInc, 61k stars. Python framework for orchestrating role-based autonomous agents
- mempalace by MemPalace, 61k stars. Open-source memory system for AI assistants, built on ChromaDB and MCP
- ai-agent-book by bojieli, 58k stars. Open-source book on AI agent design with full text, PDF, and chapter code
- multica by multica-ai, 56k stars. Platform where humans and AI agents work on tasks as one team, self-hostable
- awesome-openclaw-skills by VoltAgent, 54k stars. Filtered and categorized list of over 5,400 OpenClaw skills
- shannon by KeygraphHQ, 52k stars. AI pentesting tool that analyzes source code and runs exploits against web apps
- agent-browser by vercel-labs, 51k stars. Command line tool that lets AI agents drive a browser
- openhuman by tinyhumansai, 43k stars. Lightweight open-source agent harness meant to run many agents on a small VPS
- OpenViking by volcengine, 42k stars. Context database unifying agent memory, knowledge RAG and skills
- mindshub by mindsdb, 40k stars. Workspace where open-source models run tasks on your behalf
- skills by vercel-labs, 39k stars. CLI to find and install agent skills
- nuwa-skill by alchaincyf, 38k stars. Agent skill that extracts a person's thinking patterns into a reusable prompt
- QwenPaw by agentscope-ai, 38k stars. Personal AI assistant you deploy on your own machine or cloud, linked to chat apps
- buzz by block, 37k stars. Communication platform for coordinating a group of agents in one place
- dsh-desktop by anywhere-labs, 36k stars. Desktop client for the DeepSeek Harness plugin system
- agentscope by agentscope-ai, 36k stars. Python framework for building multi-agent apps with traceable, visible behavior
- zeroclaw by zeroclaw-labs, 33k stars. Small, fast autonomous personal assistant runtime in Rust
- cognee by topoteretes, 33k stars. Open-source memory layer giving agents persistent long-term memory
- gbrain by garrytan, 33k stars. Personal memory layer for OpenClaw and Hermes agents
New repos every two days, stars refreshed daily. Visits are counted without cookies, and the numbers are public. How it works Data RSS Newsletter