AIエージェント
Agon
Autonomous research loops, from topic to experiments.
★ 54 スターMIT
ガイドを見る →OpenScienceは、synthetic-sciencesが公開しているオープンソースのAIエージェントプロジェクトです（GitHub: synthetic-sciences/openscience）。GitHubスターは4,004、フォークは526で、Apache-2.0ライセンスの下で公開されています。GitHubでの最終更新は2026年10月13日です。
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
Install globally from npm, or run the desktop app or browser workspace if you prefer a GUI.
Describe the research task in plain language — a claim to reproduce, a dataset to clean, a parameter sweep to run.
Follow the visible trace of searches, code and file changes, then read the written-up results and check them like any lab notebook.
A turn reads as what happened — reasoning, searches, runs and writes — so nothing happens that you cannot inspect afterwards.
Shell access, Python and R kernels, notebooks, and a file system with explicit read and write grants.
Desktop app, browser workspace or terminal command — on your machine, against your files.
The desktop app updates itself; the CLI upgrades with a single openscience upgrade command.
The CLI package is published as @synsci/openscience.
npm install -g @synsci/opensciencenpx launches the workbench directly.
npx synsciKeep an npm installation current with the built-in upgrade command.
openscience upgradesynthetic-sciences; the project site is openscience.sh.
Yes, Apache-2.0 licensed and published on npm.
The workbench is free; the underlying model usage follows whichever provider account or key you configure.
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.
Run npm install -g @synsci/openscience (or npx synsci), open the workbench, and describe your research goal in plain language.
Autonomous research loops, from topic to experiments.
An autonomous research agent that plans, searches and writes cited reports from any data source
Devika - the open-source agentic software engineer that plans, browses and writes code
A lean, HuggingFace-native RL framework for agentic research.
Deep research that stays on your machine — local or cloud LLMs, 10+ search engines, encrypted and private.
An embeddable, portable, branchable microVM that runs AI agents safely on your own machine.
LibreAIカタログに掲載されたAIエージェントプロジェクトの中で最もスターの多いもの — OpenScienceが合わなかった場合に検討したい、現実的な候補リストです。
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
The agent that grows with you — a Python framework for adaptive AI agents.
DeepSeek's plugin-based agent harness where everything is a plugin.
カタログの中でOpenScienceとスター数が近いAIエージェントプロジェクト — 前後の棚もそのまま見て回れます。
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 lets AI agents drive your real Chrome — your logins, extensions, and cookies — via Playwright
Langroid: a principled Python framework for building LLM multi-agent applications (from CMU/UW researchers)
Team chat where AI agents are first-class teammates, not bots.
Basic Memory: local-first MCP memory so you and your AI tools build a shared knowledge graph in Markdown
Code, build and evaluate LLM agents with full MCP, Skills, ACP and A2A support.