scientific-agent-skills is a collection of 158 ready-to-use scientific and research skills plus access to 100+ scientific databases designed to make AI agents more capable and reliable for multi-step scientific workflows. It covers domains including biology, chemistry, medicine, and drug discovery.
Project overview
By providing 158 curated skills built on the open Agent Skills standard, the project gives coding agents version-aware workflows for scientific packages and deterministic access to scientific databases across specialized domains.
Project type
AI Agent
Use cases
Search & Research · Data Analysis
Deployment
Refer to project documentation
License
MIT
Best for
Researchers who rely on AI coding agents for scientific workflows and need curated, version-aware guidance for specialized libraries and databases.
Developers building agent-driven pipelines that require provenance-rich database lookups and integration with scientific platforms.
Key capabilities
A unified database-lookup skill providing deterministic, provenance-rich access to 78 public databases plus dedicated data access skills and multi-database packages.
Explicitly defined, version-aware workflows for scientific packages including RDKit, Scanpy, PyTorch Lightning, scikit-learn, and others.
Explicitly defined skills for Benchling, DNAnexus, LatchBio, OMERO, Protocols.io, Open Notebook, Ginkgo Cloud Lab, LabArchives, and Opentrons.
Literature review, evidence-traceable scientific writing, confidential peer review, document processing, Paperclip, and more.
Evidence-bounded hypothesis generation, grant writing, aggregate clinical decision-support research, clinician-authored treatment-plan formatting, and PK/PD modelling and simulation.
Limitations and risks
Database skills require internet access to query the relevant APIs.
Each skill has its own license specified in its SKILL.md file, which may differ from the repository's MIT License.
Skills can execute code and influence a coding agent's behavior, requiring review before installation.
As a small team, the maintainers cannot guarantee that every community-contributed skill has been exhaustively reviewed for all possible risks.
Getting started
Install a supported agent host or CLI such as npx or GitHub CLI. Then run 'npx skills add K-Dense-AI/scientific-agent-skills' or copy skills directly into your skills directory.
Evidence and sources
GitHub project description: 158 ready-to-use skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery.
README: This repository provides **158 scientific and research skills** organized into the following categories:
README: While the agent can use any Python package or API on its own, these explicitly defined skills provide curated documentation and examples that make it significantly stronger and mo…
README: A comprehensive collection of **158 ready-to-use scientific and research skills**
README: - **Save Days of Work** - Skip API documentation research and integration setup