Agents IA
Agon
Autonomous research loops, from topic to experiments.
★ 54 étoilesMIT
Voir le guide →// agents · synthetic-sciences/openscience
OpenScience
OpenScience est un projet de agents IA open source, créé par synthetic-sciences (synthetic-sciences/openscience sur GitHub). Il compte 4 004 étoiles sur GitHub et 526 forks, il est publié sous la licence Apache-2.0 et a été mis à jour pour la dernière fois sur GitHub le 13 octobre 2026.
OpenScience — the open-source AI workbench for scientific research: give it a goal, it reads the literature, runs the experiments and writes up the findings
★ 4,0 k étoiles⑂ 526 forksApache-2.0mis à jour oct. 2026
▦ visuel — le README de ce projet ne contient aucune capture intégrable,
voici donc un emplacement réservé. Le guide ci-dessous est complet malgré tout.
voici donc un emplacement réservé. Le guide ci-dessous est complet malgré tout.
// détails du projet
Détails du projet
- Étoiles GitHub
- 4 004
- Forks
- 526
- Licence
- Apache-2.0
- Dernière mise à jour sur GitHub
- 13 octobre 2026
- Catégorie
- Agents IA
- Type de fiche
- Guide d'installation complet
- Site officiel
- Official site
// fonctionnement
Comment ça marche
- Install the CLI
Install globally from npm, or run the desktop app or browser workspace if you prefer a GUI.
- Give it a goal
Describe the research task in plain language — a claim to reproduce, a dataset to clean, a parameter sweep to run.
- Review the record
Follow the visible trace of searches, code and file changes, then read the written-up results and check them like any lab notebook.
// fonctionnalités
Pourquoi on l'adore
Every step is visible
A turn reads as what happened — reasoning, searches, runs and writes — so nothing happens that you cannot inspect afterwards.
Real tools, real files
Shell access, Python and R kernels, notebooks, and a file system with explicit read and write grants.
Three ways to run it
Desktop app, browser workspace or terminal command — on your machine, against your files.
Self-updating
The desktop app updates itself; the CLI upgrades with a single openscience upgrade command.
Advertisement
// lancez-le
Lancez-le vous-même
- Install from npm
The CLI package is published as @synsci/openscience.
npm install -g @synsci/openscience - Or run without installing
npx launches the workbench directly.
npx synsci - Upgrade later
Keep an npm installation current with the built-in upgrade command.
openscience upgrade
// cas d'usage
Ce que vous pouvez créer
- Reproducing a published claim before building on it
- Cleaning and exploring a messy dataset with a full audit trail
- Running parameter sweeps and summarizing the outcomes
- Drafting methods sections and checking references against sources
Advertisement
// faq
Questions fréquentes
Who maintains OpenScience?
synthetic-sciences; the project site is openscience.sh.
Is it free?
Yes, Apache-2.0 licensed and published on npm.
Does it need a paid API key?
The workbench is free; the underlying model usage follows whichever provider account or key you configure.
How does it compare to a well-known alternative?
Hosted deep-research tools return an answer; OpenScience returns an inspectable record — the searches, code and files behind the answer — and runs locally against your own files.
How do I get started?
Run npm install -g @synsci/openscience (or npx synsci), open the workbench, and describe your research goal in plain language.
// continuez l'exploration
Outils similaires
Agents IA
gpt-researcher
An autonomous research agent that plans, searches and writes cited reports from any data source
★ 30,0 k étoilesApache-2.0
Voir le guide →Agents IA
devika
Devika - the open-source agentic software engineer that plans, browses and writes code
★ 19,6 k étoilesMIT
Voir le guide →Agents IA
labs-molt
A lean, HuggingFace-native RL framework for agentic research.
★ 1,2 k étoilesApache-2.0
Voir le guide →Agents IA
openwiki
A CLI that writes and maintains agent-ready docs for your codebase.
★ 17,0 k étoilesMIT
Voir le guide →Agents IA
local-deep-research
Deep research that stays on your machine — local or cloud LLMs, 10+ search engines, encrypted and private.
★ 9,2 k étoilesMIT
Voir le guide →// alternatives
Alternatives en Agents IA
Les autres projets de la catégorie Agents IA avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si OpenScience ne vous convient pas.
- openclaw★ 391,6 k
A personal AI that really does things — on any OS, with your data staying yours.
- superpowers★ 294,9 k
An agentic skills framework & software development methodology that works.
- ECC★ 275,2 k
Agent harness optimization system with skills, memory and security for Claude Code, Codex and more
- hermes-agent★ 252,0 k
The agent that grows with you — a Python framework for adaptive AI agents.
- deepseek-harness★ 245,5 k
DeepSeek's plugin-based agent harness where everything is a plugin.
// plus en agents
Plus en Agents IA
Projets de la catégorie Agents IA proches dans le catalogue, les plus proches en étoiles de OpenScience — continuez à explorer de part et d'autre.
- Rakazo★ 4,1 k
Rakazo — an open-source, self-hosted AI agent for persistent teammates: bring your own model and computer, run it on web, desktop, iPhone and Android
- playwriter★ 4,0 k
Playwriter lets AI agents drive your real Chrome — your logins, extensions, and cookies — via Playwright
- langroid★ 4,1 k
Langroid: a principled Python framework for building LLM multi-agent applications (from CMU/UW researchers)
- cumora★ 3,9 k
Team chat where AI agents are first-class teammates, not bots.
- basic-memory★ 4,1 k
Basic Memory: local-first MCP memory so you and your AI tools build a shared knowledge graph in Markdown
- fast-agent★ 3,9 k
Code, build and evaluate LLM agents with full MCP, Skills, ACP and A2A support.
Obtenir OpenScience
Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.