Scientific Agent Skills

Active
GitHub Python MIT

Description

A set of ready-to-use Agent Skills for research, science, engineering, analysis, finance and writing across multiple coding agents.

Key Features

  • 142 ready-to-use scientific skills — Covers 17+ fields: bioinformatics, cheminformatics, proteomics, clinical research, medical imaging, and more
  • 100+ scientific database integration — Unified access to 78+ public databases including PubChem, ChEMBL, UniProt, COSMIC, ClinicalTrials.gov
  • 70+ optimized Python package skills — Dedicated skills for RDKit, Scanpy, PyTorch Lightning, scikit-learn with curated documentation and examples
  • Agent Skills standard compatible — Works with Cursor, Claude Code, Codex, Google Antigravity and any AI agent supporting the Agent Skills standard
  • Hugging Science integration — Discover datasets, models, and demos across 17 scientific domains on Hugging Face
  • Multi-modal workflow support — Full research workflows from literature review to experimental design, including 9 scientific platform integration skills

Use Cases

💡 Drug discovery & design — Accelerate drug development with molecular docking, ADMET analysis, virtual screening skills
💡 Genomics & single-cell analysis — Perform RNA velocity analysis, gene regulatory network inference, variant annotation
💡 Clinical trial data analysis — Pull trial data from ClinicalTrials.gov for statistical analysis and visualization
💡 Scientific literature review — Automate literature search, citation management, and review writing workflows
💡 Lab automation — Integrate with Benchling, DNAnexus platforms for liquid handling and workflow automation

Quick Start

# Clone and install
export OPENAI_API_KEY={your-key}
cd scientific-agent-skills
pip install -e .

# Use in Cursor
# Search for 'PubChem' or 'molecular docking' in the Agent Skills panel
# Select and use the skill directly

# K-Dense BYOK desktop app
git clone https://github.com/K-Dense-AI/k-dense-byok
cd k-dense-byok
bun install && bun start

Related Projects