Paper2Agent (Nature): turns papers into MCP tool agents for ~$14 and 45 minutes
Justgototheeffinmoon · reddit · 2026-09-24
A Nature paper introduces Paper2Agent (Jiacheng Miao, James Zou et al.), a framework that converts research papers into runnable AI agents by packaging each paper's code, data, and methods as an MCP server a chat assistant can call directly.
Key results:
- On Google DeepMind's AlphaGenome, it generated 22 validated tools in 45 minutes at $14 compute cost, scoring 98.7±1.3% on tutorial-derived queries vs 82.7% for a baseline giving Claude direct repo access.
- Across 100 computational biology papers, it produced usable agents for 74, scoring 91.2±1.6% on 300 benchmark questions.
- One case study's assembled "AI co-scientist" identified a new splicing variant associated with ADHD risk.
About a quarter of tested papers failed due to missing code, environment issues, or non-generalizable scripts. The authors propose journals add an "agent availability" clause alongside existing data/code requirements.
Related event: Stanford's Paper2Agent Turns Papers into AI Agents, Published in Nature(2 posts)→
More from coding & agent
- KAIST's WikiTrace: evidence-linked memory fixes agent "memory-induced degradation" — dair_ai · 2026-11-11
- Why inference speed will be AI products' biggest competitive edge in 1–2 years — itsOmSarraf_ · 2026-11-11
- What happened when I moved vector retrieval into Snowflake SQL instead of orchestrating RAG outside — dima806_dima · 2026-11-11
- OpenAI engineer runs Codex for 26 days straight hunting Atlantis, spawning "vibe archaeology" — Polymarket · 2026-11-11
- He replaced a $5,000/month ghostwriter with 11 Claude Skills and earned $47,000 from one post — aryanXmahajan · 2026-11-11
- Refero Styles ships 2,000+ DESIGN.md files to give Claude Code real design taste — bbssppllvv · 2026-11-11