
Open-source agent harness from DeepSeek, built on a plugin architecture
- Open source
- Open-source agent
- 282K GitHub stars
- Multi-agent frameworks

Open-source agent harness from DeepSeek, built on a plugin architecture

Open-source super agent harness that researches, codes and creates with sandboxes

Agent meta-harness that adds swarms, memory and MCP tools to Claude Code and Codex

Open-source personal agent with long-term memory that works across chat channels

Personal agent workstation you run locally or in the cloud, reachable from many chat apps

Python agent framework with tool review, sandboxing and multi-tenant serving

Multi-agent public opinion analysis system that mines social media and writes reports

Open-source Python framework for building production RAG pipelines and agents

Open-source TypeScript framework and personal agent for autonomous AI agents

Rust agent operating system with scheduled autonomous Hands and a WASM sandbox

Multi-agent system modelled on an imperial court, with review steps and a dashboard

Meta-harness that runs, governs and shares sessions across coding agents

Self-hosted crew of AI agents on an org chart with shared graph memory

100-line minimalist LLM framework built on a graph and shared store

Open-source runtime where a queen agent builds colonies of parallel workers

Java agent framework built on Spring AI with DAG graphs and a visual studio

Run a pixel-art office of coding agents on subscriptions you already pay for

Strands toolkit with a ready agent harness, sandbox and evals in Python and TypeScript

Self-hosted workspace where task agents run on all your own machines

Python framework for building autonomous and traditional AI agents

Java framework for distributed, long-running agents with sandboxes and memory

TypeScript multi-agent runtime with durable approvals and verifiable run records

Self-hosted cloud agent cluster for teams of collaborating AI agents

Open-source Python framework for building and orchestrating multi-agent AI systems

Java framework for streaming one agent to browsers over WebSocket, SSE or WebTransport

Python framework where LLM agents wrap tasks and delegate work to sub-agents

Loop that lets Claude Code or Codex CLI run a 14-role AI company on its own

Go framework for multi-agent apps with chain, parallel, cycle and graph agents

Single Python interface for building agents across seven agent frameworks

Open-source agent operating system for building teams of agents and workflows
See every kind of AI for multi-agent frameworks: tools, agents, MCP servers and skills.Ranked: the best AI for multi-agent frameworks.