irm https://feynman.is/install.ps1 | iex The one-line installer fetches the latest tagged release. To pin a version, pass it explicitly, for example curl -fsSL https://feynman.is/install | bash -s -- 0.2.35. The installer downloads a standalone native bundle with its own Node.js runtime.
The open source AI research agent.
Installation
macOS / Linux:
curl -fsSL https://feynman.is/install | bash
Windows (PowerShell):
irm https://feynman.is/install.ps1 | iex
The one-line installer fetches the latest tagged release. To pin a version, pass it explicitly, for example curl -fsSL https://feynman.is/install | bash -s -- 0.2.35.
The installer downloads a standalone native bundle with its own Node.js runtime.
To upgrade the standalone app later, rerun the installer. feynman update only refreshes installed Pi packages inside Feynman's environment; it does not replace the standalone runtime bundle itself.
To uninstall the standalone app, remove the launcher and runtime bundle, then optionally remove ~/.feynman if you also want to delete settings, sessions, and installed package state. If you also want to delete alphaXiv login state, remove ~/.ahub. See the installation guide for platform-specific paths.
Local models are supported through the setup flow. For LM Studio, run feynman setup, choose LM Studio, and keep the default http://localhost:1234/v1 unless you changed the server port. For LiteLLM, choose LiteLLM Proxy and keep the default http://localhost:4000/v1. For Ollama or vLLM, choose Custom provider (baseUrl + API key), use openai-completions, and point it at the local /v1 endpoint.
Skills Only
If you want just the research skills without the full terminal app:
macOS / Linux:
curl -fsSL https://feynman.is/install-skills | bash
Windows (PowerShell):
irm https://feynman.is/install-skills.ps1 | iex
That installs the skill library into ~/.codex/skills/feynman for Codex. You can also name the Codex target explicitly:
macOS / Linux:
curl -fsSL https://feynman.is/install-skills | bash -s -- --codex
Windows (PowerShell):
& ([scriptblock]::Create((irm https://feynman.is/install-skills.ps1))) -Scope Codex
For a repo-local Claude/agent install instead:
macOS / Linux:
curl -fsSL https://feynman.is/install-skills | bash -s -- --repo
Windows (PowerShell):
& ([scriptblock]::Create((irm https://feynman.is/install-skills.ps1))) -Scope Repo
That installs into .agents/skills/feynman under the current repository.
For an OpenCode project-local install instead:
macOS / Linux:
curl -fsSL https://feynman.is/install-skills | bash -s -- --opencode
Windows (PowerShell):
& ([scriptblock]::Create((irm https://feynman.is/install-skills.ps1))) -Scope OpenCode
That installs into .opencode/skills/feynman under the current repository.
These installers download the bundled skills/ and prompts/ trees plus the repo guidance files referenced by those skills. They do not install the Feynman terminal, bundled Node runtime, auth storage, or Pi packages.
What you type → what happens
$ feynman "what do we know about scaling laws"
→ Searches papers and web, produces a cited research brief
$ feynman deepresearch "mechanistic interpretability"
→ Multi-agent investigation with parallel researchers, synthesis, verification
$ feynman lit "RLHF alternatives"
→ Literature review with consensus, disagreements, open questions
$ feynman audit 2401.12345
→ Compares paper claims against the public codebase
$ feynman replicate "chain-of-thought improves math"
→ Replicates experiments on local or cloud GPUs
$ feynman recipe "fine-tune a small model for math reasoning"
→ Finds ranked, implementable ML training recipes from papers, datasets, docs, and code
Workflows
Ask naturally or use slash commands as shortcuts.
| Command | What it does |
|---|---|
/deepresearch <topic> | Source-heavy multi-agent investigation |
/lit <topic> | Literature review from paper search and primary sources |
/review <artifact> | Simulated peer review with severity and revision plan |
/audit <item> | Paper vs. codebase mismatch audit |
/replicate <paper> | Replicate experiments on local or cloud GPUs |
/recipe <task-or-paper> | Ranked ML training recipes with dataset, method, code, and verification status |
/compare <topic> | Source comparison matrix |
/draft <topic> | Paper-style draft from research findings |
/autoresearch <idea> | Autonomous experiment loop |
/watch <topic> | Recurring research watch |
/outputs | Browse all research artifacts |
Agents
Four bundled research agents, dispatched automatically.
- Researcher — gather evidence across papers, web, repos, docs
- Reviewer — simulated peer review with severity-graded feedback
- Writer — structured drafts from research notes
- Verifier — inline citations, source URL verification, dead link cleanup
Skills & Tools
- AlphaXiv — paper search, Q&A, code reading, annotations (via
alphaCLI) - Hugging Face Hub — dataset metadata, split/schema inspection, and small file reads from model, dataset, and Space repos
- Docker — isolated container execution for safe experiments on your machine
- Web search — Exa, Perplexity, or Gemini API; no Chromium cookie access by default
- Session search — indexed recall across prior research sessions
- Preview — browser and PDF export of generated artifacts
- Modal — serverless GPU compute for burst training and inference
- RunPod — persistent GPU pods with SSH access for long-running experiments
How it works
Built on Pi for the agent runtime, alphaXiv for paper search and analysis, and CLI tools for compute and execution. Runtime resources follow Pi's documented package model for packages, extensions, and skills. Hugging Face inspection uses the public Hub API endpoints and HF_TOKEN / HUGGINGFACE_HUB_TOKEN environment variables documented by huggingface_hub. The ML recipe workflow was informed by the open-source Hugging Face ml-intern research-agent repo, but is implemented as native Feynman prompts, skills, and read-only tools. Every output is source-grounded — claims link to papers, docs, or repos with direct URLs.
Star History
Contributing
See CONTRIBUTING.md for the full contributor guide.
git clone https://github.com/companion-inc/feynman.git
cd feynman
nvm use || nvm install
npm install
npm test
npm run typecheck
npm run build
Videos
How to Learn Faster with the Feynman Technique (Example Included)
Thomas Frank · 7331K views
The Feynman Technique
Sprouts · 7602K views
Richard Feynman talks about the O ring
Amalek61 · 463K views
What people are saying
“I'm unsure of what is surprising to Feynman. Is he saying that he assumed clever men would be obnoxious and self-righteous when they weren't? Although an anecdote, I would imagine there are plenty of examples where the o…”
HNlbriner↗
“I still find this quite hard to grasp. If it's declarative (book) knowledge, then it makes sense to ask immediately. Does that count as struggling? If it is insight though, for even those armed with all the crucial facts…”
HNwhacked_new↗
“Science news this week: El Niño arrives, the Artemis III crew are revealed, a 'cold blob' expands across the Atlantic, and a forgotten note from Richard Feynman gets deciphered - Live Science — Live Science”
NEWSLive Science↗
“Scientists uncover Feynman’s formula for finding best holiday restaurant - The Guardian — The Guardian”
NEWSThe Guardian↗
“NVIDIA confirms RTX Spark roadmap with Rubin in 2027 and Rosa Feynman in 2029 - VideoCardz.com — VideoCardz.com”
NEWSVideoCardz.com↗
You might also like
Web
Recordly
Recordly is your open-source screen recorder and editor for walkthroughs, demos, product videos, and more. Accepting PRs. Recordly is a desktop app for recording and editing screen captures with motion-driven presentation tools built in. Instead of sending raw footage to a motion designer just to add zooms, cursor polish, or a styled background, Recordly handles that workflow in one place for free. Recordly can automatically emphasize activity with zoom suggestions, smooth cursor movement, add motion effects, and place the final composition inside a styled frame with wallpapers, colors, gradients, blur, padding, and shadows.
Web
obscura
Lightweight, stealthy, and built in Rust. Obscura is a headless browser engine written in Rust, built for web scraping and AI agent automation. It runs real JavaScript via V8, supports the Chrome DevTools Protocol, and acts as a drop-in replacement for headless Chrome with Puppeteer and Playwright. Designed for automation at scale, not desktop browsing.
Web
MiMo Code
MiMoCode is a terminal-native AI coding assistant. It can read and write code, run commands, manage Git, and use a persistent memory system to keep a deep understanding of your project across sessions while continuously improving itself. MiMo Auto is built in as a free-for-limited-time channel, so you can start with zero configuration. MiMoCode also supports connecting to any mainstream LLM provider API. The first launch guides you through configuration automatically. Supported options:
Web
Kami
Kami (紙, かみ) means paper in Japanese: the surface where a finished idea lands. AI can produce documents better than most humans do manually. The missing piece is not capability but constraint: without a design system, every session drifts into generic gray and inconsistent layouts. Kami fills that gap: one constraint language, nine templates, simple enough for agents to run reliably, strict enough that every output is something you actually want to ship. English and Chinese are first-class; Japanese and Korean work via best-effort CJK paths with visual QA before delivery. Part of a trilogy: Kaku (書く) writes code, Waza (技) drills habits, Kami (紙) delivers documents.