K-Dense-AI/scientific-agent-skills

What it solves

Scientific Agent Skills provides a comprehensive library of 163 ready-to-use skills that transform a general-purpose AI agent into a specialized research assistant. It eliminates the need for researchers to manually research API documentation and set up complex integrations for scientific databases, Python packages, and laboratory tools, allowing them to execute multi-step scientific workflows via simple prompts.

How it works

The project follows the open Agent Skills and Agent Plugins standards. It provides curated documentation, code examples, and best practices (contained in SKILL.md files) that guide an AI agent on how to use specific scientific tools. These skills can be installed into compatible AI agent hosts (such as Cursor, Claude Code, Codex, and Google Antigravity) which then discover and apply these instructions when performing scientific tasks.

Who it’s for

This is designed for scientists, researchers, and engineers across domains including bioinformatics, cheminformatics, proteomics, clinical research, materials science, physics, astronomy, and geospatial science who use AI coding agents to accelerate their research workflows.

Highlights

• Massive Database Access: Unified access to over 100 scientific and financial databases, including PubChem, ChEMBL, UniProt, and ClinicalTrials.gov.
• Optimized Package Guidance: Version-aware workflows for 70+ specialized Python packages like RDKit, Scanpy, and PyTorch Lightning.
• Broad Domain Coverage: Spans biology, chemistry, medicine, physics, and engineering, including specialized tools for PK/PD modelling and ISO standards-readiness.
• Integration Ready: Pre-defined paths for scientific platforms like Benchling, DNAnexus, and Opentrons.
• Research Tooling: Includes tools for evidence-traceable scientific writing, grant writing, and hypothesis generation.