Harnesses (76)
Agent runtimes, frameworks and harnesses: CLIs, SDKs and orchestration loops.
Machine-readable: /api/harnesses.json.
- AG2
Protocol-driven Python framework for cooperating agents, tool-calling loops, human-in-the-loop workflows and multi-agent orchestration.
- Agent S
Simular computer-use agent framework that uses screenshots and GUI actions to complete tasks in desktop and browser apps on macOS, Windows and Linux.
- Agenta
Open-source, self-hostable team workspace for building agents by chat, with tool permissions and approvals, sandboxed sessions, tracing and Slack/WhatsApp/Telegram access.
- AgentScope
Python agent framework with ReAct loops, tools and MCP/skills integration, context compaction, memory, permission checks and isolated execution backends.
- Agno
Framework and runtime to build, run and manage agent platforms (Python SDK plus runtime).
- AI SDK (Vercel)
Free, open-source TypeScript toolkit from Vercel for building AI apps and agents with streaming, fallbacks and multi-model support.
- Aider
Open-source AI pair programming in your terminal.
- Amp
Coding agent and development environment; run agents unsupervised in cloud orbs or locally with the CLI.
- AppAgent
Multimodal smartphone agent framework that learns app workflows from exploration or demonstrations and operates through taps and swipes.
- Auggie CLI
Augment Code's agentic coding CLI for the terminal, backed by its Context Engine, with custom slash commands, MCP, non-interactive mode for CI and GitHub Actions.
- AutoGen (deprecated)
Microsoft Research's multi-agent programming framework, now in maintenance mode; new projects are pointed to Microsoft Agent Framework.
- CAMEL
Open-source Python multi-agent framework from CAMEL-AI for role-playing agents, data generation, task automation and large-scale world simulation.
- Claude Agent SDK
Anthropic's SDK for building production agents with Claude Code as a library, in Python and TypeScript.
- Claude Code
Anthropic's agentic coding tool that reads your codebase, edits files, runs commands and integrates with dev tools; available in terminal, IDE, desktop and browser.
- Cline
Open-source autonomous coding agent available as an SDK, IDE extension or CLI, with Plan/Act modes and MCP integration.
- Cloudflare Agents SDK
SDK for stateful AI agents on Cloudflare with persistent state, real-time WebSockets, scheduled tasks and remote MCP servers.
- Codex App Server (beta)
Experimental JSON-RPC interface that powers rich Codex clients (e.g. the VS Code extension); embed Codex in your product. Replaces the removed codex mcp-server.
- Continue (deprecated)
Open-source coding agent (CLI, VS Code, JetBrains) whose repo went read-only after a final 2.0.0 release; no longer actively maintained.
- CopilotKit
Open-source SDK for agent-native apps: generative UI, shared state and human-in-the-loop for React, Angular, Vue, React Native, Slack and Teams, built on AG-UI.
- CrewAI
Python framework for orchestrating role-playing, autonomous AI agents that collaborate as crews and flows.
- Crush
Charm's terminal coding agent: multi-model (OpenAI- or Anthropic-compatible APIs), session-based, LSP-enhanced and extensible with MCP servers.
- Cursor
AI code editor and coding agent with Agent mode, Rules, Skills, MCP servers and a CLI.
- Deep Agents (LangChain)
LangChain's batteries-included agent harness for agents that plan, spawn subagents and use a file system for complex tasks.
- DeepSeek Harness (beta)
DeepSeek's open-source, plugin-based agent harness (dsh) with a local Web UI, a Python SDK and support for DeepSeek and third-party models.
- DeerFlow
ByteDance’s MIT super-agent harness (DeerFlow 2.0) with sub-agents, memory, skills, MCP and Docker sandboxes, installed from git rather than PyPI.
- Devin CLI
Cognition's local coding agent that runs in your terminal on macOS, Linux, WSL and Windows, with handoff to cloud Devin sessions.
- Devin Desktop (formerly Windsurf)
Cognition's AI IDE for Mac, Windows and Linux for managing local and cloud agents; windsurf.com now redirects to it.
- Dify
Source-available platform for agentic workflows, RAG pipelines and sandboxed agents on a visual canvas; self-host with Docker or use Dify Cloud.
- DSPy
Stanford NLP framework for programming, not prompting, language models.
- Factory Droids
Factory's AI coding agents (Droids) that automate coding, testing and deployment.
- fast-agent
CLI-first Python agent harness and framework with end-to-end MCP support (sampling, elicitation, OAuth), Agent Skills, ACP and declarative multi-agent workflows.
- Gemini CLI
Open-source AI agent from Google that brings Gemini models directly into your terminal.
- GitHub Copilot CLI
Brings GitHub Copilot's coding agent to your terminal to read, write and run code where you work.
- GitHub Copilot cloud agent
Copilot agent on GitHub.com that takes delegated tasks from issues, pull requests or chat, works in a cloud environment and opens a pull request for review.
- Google Agent Development Kit (ADK)
Open-source, code-first toolkit from Google for building, evaluating and deploying multi-agent systems.
- Google Antigravity
Google's agentic coding platform (app, IDE, CLI, IDE extensions and Python SDK) for orchestrating autonomous subagents, scheduled tasks and code verification.
- goose
Open-source, extensible AI agent (desktop app, CLI and API) that can install, execute, edit and test with any LLM.
- Griptape
Python framework for agents, sequential pipelines and parallel workflows, with tools, retrieval engines and task memory that keeps large outputs off prompts.
- Harness²
Apache-2.0 Google Research framework for recursively improving an agent’s prompts, tools, memory and skills with the model weights left fixed.
- Haystack
deepset's open-source orchestration framework for production-ready, context-engineered LLM applications and agents.
- Hermes Agent
Nous Research's self-improving open-source agent: terminal UI plus messaging gateway, agent-curated memory and skills it writes and refines from experience.
- Inngest AgentKit
TypeScript framework for multi-agent networks with shared typed state, code- or LLM-based routers, MCP tools and durable execution plus tracing on Inngest.
- Jules
Google's autonomous coding agent that plans and runs tasks against your GitHub repos asynchronously, with a REST API and the Jules Tools CLI.
- Kilo Code
Open-source AI coding agent for VS Code, JetBrains, CLI and cloud, with access to 500+ models and bring-your-own-key.
- Kiro
Agentic engineering environment that turns prompts into executable specs and builds across large codebases with parallel agents.
- LangChain
Open-source Python and TypeScript framework whose create_agent builds a configurable agent harness from a model, tools, a prompt and middleware.
- Langflow
Open-source visual builder for AI agents and workflows in Python; every flow can be served as an API or exposed as an MCP server.
- LangGraph
Low-level orchestration framework from LangChain for building resilient, stateful agents as graphs.
- LangGraph Swarm
LangChain's Python library for swarm-style multi-agent systems on LangGraph, where agents hand control to each other with handoff tools.
- Langroid
Python multi-agent framework with task orchestration, typed tool messages, retrieval agents and an MCP-to-tool adapter.
- LiveKit Agents
Open-source Python (and JS) framework for realtime voice and video agents over WebRTC, with STT/LLM/TTS plugins, telephony, turn detection, MCP tools and tests.
- LlamaIndex
Document-processing platform for AI (open-source LlamaIndex library plus hosted LlamaParse APIs that turn PDFs, scans and Office files into markdown or JSON).
- Mastra
TypeScript framework for AI agents and apps with memory, tools, MCP and observability built in.
- mcp-agent
Python framework for composing MCP-connected agents and workflows, with connection lifecycle management and optional Temporal-backed durable execution.
- Microsoft Agent Framework
Framework for building, orchestrating and deploying AI agents and multi-agent workflows in Python and .NET; successor to AutoGen.
- mini-swe-agent
Minimal open-source software-engineering agent from the SWE-bench/SWE-agent team: about 100 lines of Python, bash as its only tool, a linear history, any model via LiteLLM.
- n8n
Fair-code workflow automation platform for AI agents with 1,500+ integrations, an MCP Client node, MCP Server Trigger and a built-in instance-level MCP server.
- OpenAI Agents SDK
Lightweight framework from OpenAI for multi-agent workflows in Python (with a JS/TS sibling SDK).
- OpenAI Codex CLI
OpenAI's lightweight open-source coding agent that runs in your terminal.
- OpenClaw
Open-source personal AI assistant that runs on your own machine behind one Gateway and talks to you over WhatsApp, Telegram, Slack, Discord, iMessage and 20+ other channels.
- OpenCode
The open-source AI coding agent, built for the terminal.
- OpenHands
Open-source, model-agnostic platform for cloud coding agents.
- Pi
Minimal, extensible coding-agent harness and TypeScript toolkit (unified LLM API, agent core, TUI) from Earendil; customize with extensions, skills and prompt templates.
- Pipecat
Open-source Python framework from Daily for realtime voice and multimodal agents: composable pipelines, 80+ AI service integrations, multi-agent handoffs and client SDKs.
- Pydantic AI
Typed Python agent framework from the Pydantic team covering agents, realtime voice, image generation and embeddings across many models.
- Qwen Code
Open-source AI coding agent for the terminal from the Qwen team.
- Roo Code (deprecated)
Shut down May 15, 2026: open-source VS Code coding agent (a Cline fork) with Code, Architect, Ask, Debug and custom modes; repo archived.
- SandBase Harness
Local-first Apache-2.0 agent runtime with sandboxed sessions, memory, credentials, MCP toolsets, audit/replay and a local Console.
- Semantic Kernel
Microsoft's open-source SDK for integrating LLM capabilities quickly and easily into your apps.
- smolagents
Hugging Face's barebones Python library for agents that think in code.
- Strands Agents
Open-source, model-driven SDK and ready-made harness for building AI agents in Python and TypeScript, with MCP, multi-agent patterns, memory, guardrails, tracing and evals.
- SWE-agent
Princeton/Stanford research agent that lets an LM fix real GitHub issues, solve CTFs or run custom tasks, configured by one YAML file.
- TapeAgents
Agent framework built around structured replayable session tapes for debugging, resuming, serving and improving single agents or multi-agent teams.
- VS Code Agents (agent mode)
Agent mode in Visual Studio Code: an agent finds relevant code, edits across files and runs checks in your workspace while you review the changes.
- Warp
Agentic development environment built from the terminal, with a built-in coding agent or your own CLI agent; client source now on GitHub (AGPL-3.0).
- Zed (Zed Agent)
Open-source, high-performance code editor with a native agent (Zed Agent), external ACP agents, terminal agent threads, MCP servers and skills.