ResearchSkills
Hejia GengResearchSkills.ai is an open platform that turns real scientific workflows into reusable skills for AI agents. Instead of storing raw chat logs, it reconstructs research sessions as decision trajectories — how researchers form hypotheses, diagnose failures, choose methods, and decide when to pivot — and distills them into portable skills that agents can retrieve and execute. Contributions are extracted locally, automatically de-identified, reviewed by domain experts for scientific accuracy, and published to an open library spanning 155+ scientific subdomains under CC BY 4.0. The goal is to give AI systems not just raw capability, but the tacit research judgment that determines which experiments matter, which dead ends to avoid, and when to persist versus change direction.
ResearchSkills.ai
The Library of Alexandria for AGI Scientists
ResearchSkills.ai is an open platform where researchers contribute, review, and discover reusable AI agent skills extracted from real scientific workflows.
Rather than storing raw chat logs, it captures decision trajectories — structured maps of research decisions such as hypothesis formulation, failure diagnosis, and method pivots — and distills them into reusable skills that any AI agent can retrieve and execute.
How it works
Researchers use a local extraction tool — such as /extract-knowhow in Claude Code or Codex — to reconstruct their research sessions as decision trajectories.
These trajectories are then:
- Automatically de-identified
- Reviewed by domain experts for scientific accuracy
- Published to an open library
The library spans 155+ scientific subdomains and is released under CC BY 4.0.
Why it matters
AI agents may have strong raw capabilities, but they still lack the tacit expertise that separates frontier researchers from those simply following steps.
That expertise includes questions like:
- Which experiments are actually worth running?
- Which dead ends should be avoided?
- When should a researcher pivot rather than persist?
ResearchSkills encodes this kind of judgment into a form that AI agents can retrieve, interpret, and execute.
Compatibility
ResearchSkills is compatible with:
- Claude Code
- Codex
- Cursor
- Windsurf
- Any AI tool that can read Markdown
Built by
Built by ScienceIntelligence, a non-profit academic collaboration.
Significant contributors are credited as co-authors on published research.