What is this?
An open source AI research agent: a terminal app that searches papers, runs literature reviews and deep-research workflows, and links every claim back to its sources.
Is it a terminal tool or does it have a graphical interface?
Both. It runs as an interactive terminal REPL where you type requests or slash commands, and feynman serve launches a local science workbench with project/session navigation, notebooks and in-app chat.
Do I need Node.js or an API key to run it?
The standalone installer bundles Node, so nothing extra; the npm route needs a local Node.js (>=22.22, <26). No fixed key: point it at local LM Studio, Ollama, LiteLLM or vLLM, or run feynman model login for hosted providers.
Can I trust its claims?
Outputs are source-grounded: every claim links to papers, docs or repos by direct URL, and feynman rank scores papers with transparent evidence for citations, methods, reproducibility and provenance.
What are the four built-in research agents?
Researcher gathers evidence across papers, web, repos and docs; Reviewer gives severity-graded critique; Writer turns research notes into structured drafts; Verifier checks inline citations, verifies source URLs and cleans up dead links. Workflow prompts invoke them when a task benefits from decomposition.
What is Feynman Bio Tools?
A Feynman-owned open science connector catalog: OpenAlex and arXiv workflows, PubMed metadata and PMC full text, bioRxiv/medRxiv preprints, ClinicalTrials.gov trials, FDA labels, ChEMBL and PubChem chemistry, Ensembl and UniProt genomics, pathways, GWAS and dozens more life-science sources.
What can I open inside the serve workbench?
Built-in previews for reports, PDFs, images, audio, video, XLSX workbooks, Jupyter notebooks, LaTeX, sequences and alignments, chemistry sketches (KET/RXN/CDXML/CXSMILES), structures, trees and tensors — plus notebooks, compute targets, cloud export and session history.
What are similar projects?
foreman, Deepgraph, Raven, ordewell, imozuru, open-science, openscience, ruflo
Which topics is it in?
literature-review, bioinformatics, deep-research, pi-coding-agent, science, arxiv, papers, opencode, local-llm, ollama, multi-agent, coding-agent