Agents IA
openclaw
A personal AI that really does things — on any OS, with your data staying yours.
★ 391,6 k étoilesMIT
Voir le guide →Agents IA autonomes et frameworks multi-agents.
Les projets les plus étoilés de cette catégorie, plus une sélection tournante de trois autres projets tirés du reste de la liste — renouvelée pour que tout le catalogue ait son tour en vitrine.
A personal AI that really does things — on any OS, with your data staying yours.
An agentic skills framework & software development methodology that works.
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
Governance and orchestration for AI agent fleets: declarative rules, verifiable runs.
Parallel multi-agent research that compiles wikis from any topic.
Orchestrate Claude Code, Kiro, and Codex together in isolated tmux sessions.
The agent engineering platform
Framework for orchestrating role-playing, autonomous AI…
A programming framework for agentic AI
LlamaIndex is the document processing platform for AI
Build resilient agents
A lightweight, powerful framework for multi-agent…
Open source agentic operating system
🐫 CAMEL: The first and the best multi-agent framework.…
Open Source framework for voice agents, multimodal…
A framework for building realtime voice AI agents 🤖🎙️📹
Open-source AI orchestration framework for building…
An agentic skills framework & software development…
🌟 The Multi-Agent Framework: First AI Software Company,…
Write HTML. Render video. Built for agents.
Ultra-lightweight, open-source, self-hosted personal AI…
AI Agent Assistant & development framework that…
817 structured cybersecurity skills for AI agents ·…
Mastra is the modern TypeScript framework for…
TencentDB Agent Memory is a team-level memory hub for…
An Open Phone Agent Model & Framework. Unlocking the AI…
A framework for building agents and services
Agent Zero AI framework
AI 时代的伯克希尔：基于 Claude Code / Codex…
Python SDK for Agent AI Observability, Monitoring and…
A framework for building, orchestrating and deploying…
The elegant testing framework for PHP developers and AI…
Open-source framework for conversational voice AI agents
Agentic AI Framework for Java Developers
AI Agent Engineering Platform built on an Open Source…
The fullstack MCP framework to develop MCP Apps for…
Omnigent is an open-source AI agent framework and…
PraisonAI 🦞 — Hire a 24/7 AI Workforce. Stop writing…
Flexible and powerful framework for managing multiple…
The Enterprise-Grade Multi-Agent Orchestration…
Product Management skills framework built on…
A framework for building agentic apps
Open-source framework for building agentic apps in…
Python SDK for AI agent monitoring, LLM cost tracking,…
A visual playground for agentic workflows: Iterate over…
Enterprise-grade AI agent framework with multi-provider…
AI agent framework for plan-first development workflows…
Koog is a JVM (Java and Kotlin) framework for building…
Agent framework for the JVM. Pronounced Em-BAY-bel…
The Ruby-native AI framework. Chats, agents, tools,…
AdalFlow: The library to build & auto-optimize LLM…
Open-source, vision-first browser agent
A股多Agent投研框架 —…
Framework for AI agents to build and maintain a digital…
Build production-ready AI agents in both Python and…
f.k.a. Awesome ChatGPT Prompts. Share, discover, and…
FULL Augment Code, Claude Code, Cluely, CodeBuddy,…
100+ AI Agents, Agent Skills and RAG Apps - Free and…
Open-source web crawler and scraper for LLMs and AI…
An open-source long-horizon SuperAgent harness that…
Bring projects, wikis, and teams together with AI.…
Open-source AI job search agent and job finder: scan…
Open-source AI penetration testing tool to find and fix…
The best-benchmarked open-source AI memory system. And…
The open-source AI voice studio. Clone, dictate, create.
an open source, extensible AI agent that goes beyond…
Open-Source Frontier Voice AI
Make humans and AI agents work as one team —…
LocalAI is the open-source AI engine. Run any model -…
An open-source, privacy-first, self-hosted knowledge…
A one-of-a-kind resume builder that keeps your privacy…
Open-source foundation of ToolJet AI - the enterprise…
Open-source red-teaming framework that stress-tests…
Native control plane for Claude Code, Codex, Cursor,…
Framework-agnostic, git-native standard for defining…
Write coding-agent rules once, then apply them to every…
Cross-platform desktop panel for managing OpenClaw and…
Blazing-fast business intelligence built for humans and…
Living architecture documentation that both your team…
Self-hosted AI pentesting agent that finds flaws and…
Battle-tested coding rules for AI agents, distilled…
Autonomous AI trading agent spanning 1000+ markets,…
Chinese-language course on AI basics: GPU architecture,…
Out-of-the-box open-source AI agent for desktop and…
Real-time voice runtime that keeps AI agents talking,…
Open-source observability for AI agents: trace,…
Model-agnostic marketplace of agent skills with a…
Turn papers, text, or topics into editable research…
Terminal security that stops malicious prompts, URLs,…
Rebuild Claude Code’s core in ~5,000 lines: an…
Chinese Java course: build AI agents with Spring AI,…
Persistent memory layer for AI agents, backed by…
Sync your AI coding setup: skills, agents, rules, MCP,…
Multi-agent court inspired by ancient China:…
Official agent skills that teach AI coding tools to…
YAO: rigorous engineering, evaluation, and governance…
Free Chinese AIGC course: prompting, agents, image,…
Rust terminal multiplexer with a typed SDK: drive any…
Multi-agent platform where every agent gets its own…
AI-native search that indexes code, docs, and PDFs with…
Chat with your data: an LLM agent that builds charts…
Build, run, and scale AI agents like APIs and…
Make your agents learn from experience with a dedicated…
Modular Python framework for AI agents and workflows…
Chinese guide to building production-grade AI agents…
Offline multi-agent simulation and prediction engine:…
Apache framework for apps that make decisions: agents,…
Simple desktop AI assistant with knowledge bases,…
Open skill that turns novels and stories into…
Community-built terminal coding agent: bring your own…
180 free gradients for CSS, Figma, and Sketch, plus a…
Hook and skill guard that stops coding agents from…
Sub-millisecond VM sandboxes for AI agents via…
AI full-stack engineer agent built on Next.js, Claude,…
Vercel’s SDK for durable, reliable, observable apps and…
Bring Claude Code, Codex, and Gemini CLI into Obsidian…
Use LLMs as plain Python functions: prompts become…
Open-source AI assistant framework with skills and an…
A markup-driven CSS orchestration language for…
Serverless AI agent framework running on…
Autonomous agent framework for Elixir: distributed,…
AI-plus-IoT framework bringing AI agents to ESP32 and…
Go framework for production agent systems: graphs,…
Android GUI agent framework: AI that sees, plans, and…
Retired agent framework: its prompts and guides live on…
Autonomous data-labeling agents: let AI do the…
Agentic AI framework for automating enterprise…
AI-native visual analytics framework that helps agents…
Let Claude Code, Codex, Gemini, and Copilot message…
Full-stack AI agent for coding, research, content, and…
JavaScript framework for multi-agent systems with a…
Build AI agents in a no-code visual builder or…
Agentic framework for long-horizon autonomous…
Governance and orchestration for AI agent fleets:…
Parallel multi-agent research that compiles wikis from…
Orchestrate Claude Code, Kiro, and Codex together in…
Up to 20x faster AI coding with smart model routing and…
Manage parallel AI coding agents on an infinite…
A code-first Java toolkit for building AI agents.
Turn a fresh Ubuntu VPS into a multi-agent dev…
Local-first AI chat workspace for models.
Dev agent multiplexer designed for git worktrees and…
Lean coding agent written in Rust, optimized for memory…
Curated skills, sub-agents, and config templates that…
Open source ETL/ELT you deploy on your own servers or…
AI Assistant that’s easy to setup.
Management console for installing and monitoring AI…
Self-improving memory layer that works across AI agent…
A coding agent that runs on your model, your machine,…
Run AI agents on desktop apps and CLI for hours without…
Official MathWorks MCP server: drive MATLAB from AI…
Context layer so AI agents query your analytics…
A memory operating system for long-term, personalized…
Pre-execution guard that stops AI agents' destructive…
Apache RocketMQ rewritten in Rust for the AI era.
Observability that dispatches AI agents to self-heal…
Compression gateway that cuts AI coding agents' token…
Self-hosted SERP API: browser-rendered search from six…
Google's code-first TypeScript toolkit for building AI…
Shared rooms where humans and cross-vendor AI agents…
C99+ style guide with an AI agent skill for formatting…
Code Editor for the AI Agents Era.
Team memory for engineers and their AI agents.
Self hosted, live digital human agent system.
Zig code intelligence server and MCP toolset for AI…
Connect your browser to AI models with a Dia-style…
A markdown knowledge graph with LSP for editors and MCP…
Give your AI coding tool expert Swift Concurrency…
Run chops locally: your ai agent skills, finally…
Gortex, in Go — built for high-performance…
Self-hosted ai agents: minimal ai coding agent (~1,000…
Run context-engineering-kit locally: hand-crafted…
Open ai agents for spec-driven development with…
Tradememory-protocol — decision audit trail +…
Nocturne_memory, in Python — built for a lightweight,…
Run poco-claw locally: a more beautiful and…
Open ai agents for mechanical verification for ai…
A personal AI that really does things — on any OS, with…
Turn any website into clean, agent-ready data with one…
A TypeScript toolkit for building agents: unified LLM…
A coding agent tuned for open models like Kimi, GLM,…
Agent memory that learns — so your agents get smarter…
Build, run, and manage full agent platforms — not just…
Program language models with code and compilers — not…
Your personal AI assistant — self-host it, chat…
Real-time knowledge graphs that give agents a memory…
The batteries-included agent harness from the LangChain…
Microsoft's SDK for weaving LLMs into real…
The open protocol that lets AI agents talk to each…
One RAG framework for everything — including multimodal…
The veteran open-source framework for conversational AI…
A CLI that writes and maintains agent-ready docs for…
Turn messy documents into clean, structured data for…
Observability and evaluation for AI apps — see what…
A blazing-fast enterprise AI gateway: 1000+ models,…
All your agents and workspaces in one place — on every…
Put guardrails on LLMs: validate outputs before they…
A payments protocol for the internet — built on plain…
Open-source AI coworkers, each with a computer of their…
Vercel's open framework for building agents that run…
Examples and integrations for Zep — the memory layer…
Team chat where AI agents are first-class teammates,…
Build voice-based LLM agents with a modular,…
A memory management kit for agents — remember me,…
Every AI agent deserves a wallet — onchain tools from…
30 runnable notebooks teaching everything about agent…
A Python framework for building MCP servers and agent…
An enterprise MCP gateway: governed, auditable tool…
A self-improving QA agent that writes tests in plain…
Scan AI models for serialization attacks before you…
Take multi-agent AutoGen workflows to production, fast.
Detect prompt injections and jailbreaks before they hit…
The tougher, more robust successor to the MMLU…
Benchmark computer-use agents on long, real-world…
A benchmark for comparing AI agent memory systems…
eBPF-powered information-flow enforcement for safer AI…
Local-first AI memory that runs offline on just 8 GB of…
Build, deploy, and monetize autonomous agents across…
The agent that grows with you — a Python framework for…
Skills and plugins that teach coding agents to write…
Give your coding agents a persistent memory that…
Convert PDFs, DOCX, and HTML into clean, AI-ready…
The cloud-native vector database for planet-scale…
Simple, fast retrieval-augmented generation with graph…
Secure, fast microVMs that make serverless possible.
A high-performance vector search engine for next-gen AI.
The memory API that gives AI apps instant, lasting…
Open-source search infrastructure built for AI…
The open-source studio for labeling data of every type.
Stateful AI agents with memory that learns and improves.
Google's code-first toolkit for building real AI agents.
Unit tests for LLMs — evaluate before you ship.
Vector search plus structured filtering in one database.
Secure sandboxes where AI agents run real code.
Scan your LLM for vulnerabilities before attackers do.
Local-first microVMs for fast, programmable sandboxes.
Production-ready agentic RAG behind a clean REST API.
OpenTelemetry-native observability for LLM apps.
Policy, identity, and sandboxing guardrails for AI…
The benchmark that grades AI on real GitHub issues.
Talk to your computer — an open-source voice interface.
A team workspace for building AI agents together.
SEO, GEO, and marketing skills your agents can use.
A rigorous benchmark for LLMs acting as agents.
An open protocol for secure AI-driven payments.
A lean, HuggingFace-native RL framework for agentic…
Official SDKs for tracing with LangSmith.
An open-source speech language model from Hume.
A memory and retrieval server built for LLMs.
Test your agents against prompt-injection attacks.
See what your AI agents do — at the system level.
A graduate-level benchmark even Google can't crack.
Guardrails for building secure, robust agents.
Safety testing for agents, right in pytest.
A complete evaluation toolkit for AI agents.
A voice-driven multi-agent assistant that orchestrates…
A personal AI assistant that puts security first.
Autonomous research loops, from topic to experiments.
DeepSeek's plugin-based agent harness where everything…
The open-source library that lets AI agents control a…
The official collection of reference Model Context…
Drop-in persistent memory infrastructure for AI agents…
A fast browser-automation CLI built for AI agents.
A self-evolving context database unifying agent memory,…
High-performance serving for the LLMs that power your…
Open-source AI memory that gives agents persistent…
Over a thousand tool integrations so agents can turn…
Apple's array framework for running ML efficiently on…
The fast, Pythonic way to build MCP servers and clients.
The official Python SDK for Model Context Protocol…
A visual all-in-one platform for building, debugging,…
A dev-first framework for building and running…
A secure, fast sandbox runtime where AI agents can…
The official TypeScript SDK for Model Context Protocol…
An open-source SDK for production agents on any model,…
Simple workflow patterns for building effective agents…
A batteries-included browser API so agents can automate…
Programmable guardrails to keep LLM systems safe and on…
Open-source LLM observability: monitor, evaluate, and…
A framework for evaluating and improving AI agents.
The open-source AgentOS, formerly known as AutoGen.
Microsoft's open framework for red-teaming generative…
Research-backed lifelong memory for LLM agents, text…
An open-source multimodal retrieval engine for agent…
NVIDIA's object-oriented, Pythonic approach to building…
A toolkit for developing AI-powered apps, from the…
Fast, best-practice evaluations for AI model outputs.
An open benchmark for browser AI agents on everyday…
A minimal harness that lets an LLM fully control a Mac.
Open-source conversational voice AI agents.
A very large curated list of AI tools, updated for 2026.
A benchmark testing whether agents can solve…
A harness-oriented framework for production-grade agent…
Agentic memory for LLM agents, from a published…
An MCP-native AI coding agent and runtime in a single…
A frozen snapshot of the original BabyAGI…
A framework for agent nexuses where intelligence…
The official Python SDK for building on Vapi's voice AI…
The SDK for extracting data and interacting with any…
An MCP server that puts Hyperbrowser's browsers behind…
Your self-hostable AI second brain for answers, agents,…
Microsoft's Playwright-powered MCP server for browser…
A benchmark for tool-agent-user interaction in…
Agent harness optimization system with skills, memory…
The open-source platform for building and running…
Multi-agent LLM framework that mirrors a trading firm:…
Give your AI agent eyes on the whole internet: Twitter,…
Open-source AI-driven development platform: autonomous…
Agent harness for multi-agent swarms: deploy,…
Chrome DevTools for coding agents: an MCP server that…
High-performance code intelligence MCP server: your…
A fast, local-first open-source agent harness written…
ByteDance's open-source multimodal AI agent stack: GUI…
A Claude Code-style coding agent that runs anywhere and…
GitHub's official MCP server — let AI tools work with…
Build and run agents you can see, understand and trust…
A lightweight, container-isolated personal AI agent for…
A Python simulation platform for robotics and embodied…
An autonomous research agent that plans, searches and…
Alibaba's in-page GUI agent: control any web interface…
Hugging Face's LeRobot: datasets, models and tools to…
A fully local Manus-style autonomous agent: browses,…
Test prompts, agents and RAGs — LLM evals, red teaming…
Autonomous AI agents for penetration testing —…
An MCP server that lets Claude, Cursor and Windsurf…
NVIDIA's reference stack for running AI agents more…
Open-source LLM observability: trace, evaluate and…
Yohei Nakajima's autonomous task agent - now a…
Stanford's Generative Agents simulation - 25 AI…
A self-improving agent for coding workflows and…
Give it a GitHub issue and it tries to fix the code for…
Kortix - the open-source AI operating system: agents,…
Tongyi Deep Research - Alibaba's open-source deep…
NVIDIA's security scanner for AI agent skills - check…
Devika - the open-source agentic software engineer that…
Microsoft's trainer for AI agents - reinforcement…
A self-healing harness that connects an LLM directly to…
Build customer-facing AI agents that actually follow…
The JavaScript/TypeScript version of LangChain - the…
An open-source agent operating system in Rust -…
A production-grade platform for building agentic chat…
Agent-native memory infrastructure - structured,…
Google's MCP Toolbox for Databases - give agents safe,…
A serverless, single-file memory layer for AI agents -…
Supercharge LLM app evaluations with objective metrics,…
Automated penetration-testing agent framework powered…
Open-source hub of integrations, SDK and CLI for…
Microsoft's R&D-Agent: LLM agents that automate…
Bridge AI assistants to the Unity Editor via MCP:…
Physical Intelligence's open vision-language-action…
Open-source agentic browser your AI agents drive with…
Production-ready headless browser for AI agents, built…
One portable, local-first memory layer for every AI…
AI-powered reverse engineering assistant that bridges…
MCP server letting AI agents run 150+ cybersecurity…
Enterprise LLM app platform: GenAI workflows, RAG,…
Self-evolving memory OS for LLMs and agents: persistent…
Local-first agentic workspace where Claude Code, Codex…
MCP server that lets AI agents run thousands of Apify…
Microsoft's GUI agent that grew into UFO³ —…
Control Android and iOS devices with natural language —…
The open runtime for AI agents — a Rust platform for…
Mobile-Agent — multimodal GUI agents that operate…
Deep research that stays on your machine — local or…
A GEP-powered self-evolving engine for AI agents —…
iFLYTEK's enterprise agentic workflow platform —…
FIRERPA / LAMDA — full-stack Android device control:…
Isaac Lab — NVIDIA's unified framework for robot…
NVIDIA Isaac GR00T N1.7 — an open foundation model for…
Stream's open Vision Agents — build real-time voice and…
Osaurus — own your AI: a native macOS agent harness…
Upsonic — build autonomous AI agents in Python, with…
Honcho — Plastic Labs' memory library for stateful…
Reef — infrastructure for continually self-improving…
Voyager — the open-ended embodied Minecraft agent that…
Open-source 7B vision-language-action model and…
Open Multi Agent — a self-hosted TypeScript agent…
AppAgent — Tencent's multimodal agent that learns…
An embeddable, portable, branchable microVM that runs…
Durable, composable AI agents — flows that crash and…
An MCP server and CLI that lets AI agents build, test…
A Large Action Model framework for building AI web…
A local-first AI agent workspace for coding, writing,…
A control plane with a durable state kernel that keeps…
An Agent OS with interview-gated, budgeted evolution —…
A desktop-grade AI agent from NetEase Youdao that does…
An open-source auth gateway connecting 1,500+ SaaS…
A zero-code platform for auto-generating…
An MCP integration platform that lets AI agents use…
A full-lifecycle platform for developing, debugging,…
An open-source platform for AI-native application…
The harness of harnesses: a persistent, self-evolving…
Connect your Zotero research library to Claude and…
An open agent framework with a native terminal TUI,…
Exa's MCP server for web search and crawling — give…
The unofficial Home Assistant MCP server — control and…
Build production-ready agentic workflows from natural…
Mobile app automation and verification built for AI…
Ghidra MCP server with 200+ tools that let AI…
Rivet is Ironclad's visual IDE and TypeScript library…
Deep Research with any LLM: in-depth reports in…
Convene 18 legendary thinkers as an AI council that…
LongMemory: local-first memory for AI agents with…
M-flow is a bio-inspired cognitive memory engine for AI…
Pydantic Logfire: Python and AI observability on…
Memory-LanceDB-Pro gives OpenClaw agents hybrid search,…
nono: zero-trust, zero-setup, zero-latency sandboxing…
Archestra: the enterprise AI platform — chat UI, LLM…
ArcBox: a fast macOS-native container and VM runtime —…
Excel MCP Server: let AI agents read, write, format,…
Basic Memory: local-first MCP memory so you and your AI…
Langroid: a principled Python framework for building…
Playwriter lets AI agents drive your real Chrome — your…
Code, build and evaluate LLM agents with full MCP,…
OpenAI's lightweight TypeScript framework for…
A low-code tool for building multi-agent LLM apps —…
A portable AI agent runtime for the JVM — write one…
Google's Python SDK for building AI agents on the…
OpenClaw rebuilt in Go — multi-tenant AI agent teams…
A token-conscious MCP server that connects AI tools to…
An open-source sandboxed agent harness for teams — a…
EvoAgentX — a framework for building self-evolving…
Control Gmail, Calendar, Docs, Sheets, Drive and more…
AI agents that use real Android and iOS apps like a…
A local MCP server for serious arXiv work — LaTeX…
OSWorld — benchmark multimodal agents on open-ended…
Snyk's security scanner for AI agents, MCP servers and…
An auto-company that works 24/7 on your own PC — 14…
Decentralized P2P mesh for autonomous AI agents — your…
User-profile-based long-term memory for AI chatbot…
Fork() for AI agent microVMs — spawn 100 KVM-isolated…
Secure persistent personal agent server in one Rust…
JADX plugin + MCP server that lets LLMs analyze Android…
Production-grade MCP server for Godot — let AI build…
Recursive self-improving physical agent OS — verified,…
MCP server, CLI and skills for searching and…
MCP aggregator, orchestrator, middleware and gateway in…
Python library for building multimodal language agents…
Self-learning browser infrastructure for AI agents —…
NVIDIA's toolkit for connecting, profiling and…
Bright Data's MCP server: web search, scraping,…
The open-source memory layer for autonomous agents —…
Long-term memory runtime for AI agents — plain Markdown…
Self-hosted MCP gateway and control plane — connect,…
Native MCP plugin for x64dbg — let an AI drive the…
Pinvou Agent — open-source desktop AI agent for tools,…
Open-source AI SOC — alert fusion, LLM triage, MITRE…
Google's modular toolkit that lets AI coding agents…
Open-source Computer Use as an MCP service: let any…
Infisical's credential vault and HTTP proxy, so AI…
A memory agent for your agent fleet: it curates,…
Self-managing, unbounded context and long-term memory…
Agno's self-learning data agent: grounded text-to-SQL…
Unla MCP Gateway: turn existing APIs and MCP servers…
Self-hosted AI vulnerability research: agents find,…
Open Cowork: a desktop AI agent app with one-click…
Shrimp Task Manager: an MCP server that turns natural…
One open platform to simulate, evaluate, guardrail,…
Uber's ADR: discover, observe, benchmark, and detect…
Gini: a personal agent runtime that remembers, learns…
Babysitter enforces the process: deterministic,…
Reliable computer use for any agent: drive real apps…
LangMem: long-term memory primitives for LangGraph…
Brave's official MCP server: web, news, image, video,…
ByteDance's SandboxFusion: a secure sandbox for running…
Qwen-AgentWorld: a language world model simulating MCP,…
SimWorld: an Unreal Engine 5 simulator for evaluating…
SWE-Gym: the first training environment for real-world…
Snowflake's Agent World Model: synthesise 1,000…
WorkArena: ServiceNow's benchmark for how well web…
WebArena-Verified: the audited, deterministically…
ZCode — Z.ai's AI coding workspace: desktop app,…
OpenScience — the open-source AI workbench for…
Rakazo — an open-source, self-hosted AI agent for…
Monid — OpenRouter for agent tools: one base URL and…
Loop Engineering — practical patterns, starters and a…
Talorys — your personal AI agent in your own Cloudflare…
Tutti — a real-time shared workspace where people and…
ESP32 iPhone Agent — let an AI model use a real iPhone…
OpenMausBot — the open-source Grok Bot alternative:…
Beam — Reflection's open 501B MoE model built for…
smolagents — Hugging Face's barebones agent library…
Jev Ultrafast — the fastest browser agent from…
Swarm — OpenAI's experimental framework for lightweight…
Autres catégories : VidéoImageAudioMusique IAIA 3DClips courtsLLM locauxChat IARAG et connaissancesAutomatisationProductivitéDéveloppementDesignÉcritureAuto-hébergementInstallation facile
Les agents IA vont au-delà du chat : ils planifient un travail en plusieurs étapes, appellent des outils, naviguent, écrivent et exécutent du code, puis rendent compte. Cette catégorie couvre les frameworks à agent unique, les équipes multi-agents où des agents spécialisés collaborent, les agents d'usage d'ordinateur qui pilotent un bureau, et les harnais qui les gardent observables et sûrs.
La ligne de partage honnête est la quantité de supervision que chaque système exige. Certains sont des frameworks de recherche pour construire votre propre agent ; d'autres sont des travailleurs prêts à l'emploi pour le code, la recherche ou les tâches de navigateur. Chaque guide d'ici dit explicitement ce que l'agent peut réellement faire aujourd'hui, car c'est la catégorie où les démos dépassent le plus souvent la réalité.
Cette page rassemble 478 projets agents IA open source sélectionnés par LibreAI — Agents IA autonomes et frameworks multi-agents. Chacun est vérifié sur GitHub et a sa propre page guide avec une présentation, des étapes d'installation, des cas d'usage et une FAQ.
Oui. Tout outil ici est open source et gratuit à télécharger et à utiliser sous sa propre licence — la plupart utilisent MIT, Apache-2.0 ou GPL-3.0, et toute licence personnalisée de type « code source disponible » est signalée sur la page de l'outil. Vérifiez le fichier LICENSE avant un usage commercial.
Non. LibreAI ne publie que des guides — aucun téléchargement n'est servi depuis ce site. Chaque installation provient du dépôt GitHub officiel du projet.
Ouvrez le guide de l'outil et suivez sa section « Lancez-le vous-même » — les étapes viennent de la documentation du projet lui-même — ou utilisez le bouton « Voir sur GitHub » pour aller directement au dépôt officiel.
Chaque projet est vérifié en direct via l'API GitHub — étoiles, forks, licence et activité récente — et n'est référencé que s'il a un véritable élan et une communauté active.
Les tâches courtes et bien définies — répondre à des questions sur un code source, rédiger des résumés de recherche — fonctionnent bien aujourd'hui. Les longues tâches ouvertes dérivent encore, c'est pourquoi les meilleurs frameworks d'ici mettent l'accent sur les traces, les points de contrôle et les approbations humaines plutôt que sur l'autonomie totale.
Traitez un agent comme un nouveau prestataire : commencez avec un accès limité, des environnements en bac à sable et la relecture des actions conséquentes. Les projets de cette catégorie diffèrent largement dans leur outillage de sécurité, et les guides signalent ceux qui ont de vraies protections.