/ Directorio / Playground / scientific-agent-skills
← Todos los servidores ↗ GitHub
● Comunidad K-Dense-AI ⚡ Instantáneo

scientific-agent-skills

por K-Dense-AI · K-Dense-AI/scientific-agent-skills

133 domain-expert skills covering bioinformatics, drug discovery, clinical data, ML, and scientific writing — Claude becomes a credible research collaborator.

A library of Agent Skills for scientific workflows. Each sub-skill (BioPython, RDKit, Scanpy, DeepChem, PyMC, PubMed lookup, etc.) ships its own SKILL.md with usage patterns and references. With the bundle installed, Claude stops guessing at bioinformatics APIs and instead follows the domain's actual idioms, tool names, and data formats.

▶ Ver demo 📦 Instalar 💡 Casos de uso

Por qué usarlo

Características clave

Demo en vivo

Cómo se ve en la práctica

scientific-agent-skill.replay ▶ listo
0/0

Instalación

Repositorio del autor · Documentación oficial de Skills de Claude

Casos de uso

Usos del mundo real: scientific-agent-skills

Run a single-cell RNA-seq analysis with Scanpy

👤 Computational biologists and bioinformatics postdocs ⏱ ~30 min intermediate

Cuándo usarlo: You have an .h5ad file and want QC, clustering, and UMAP without re-reading the Scanpy docs.

Requisitos previos
  • Scanpy installed in your env — pip install scanpy
  • An .h5ad dataset — Download from a public repo or use your own
Flujo
  1. Tell Claude the data and the goal
    I have pbmc3k.h5ad. Use the scanpy skill to do QC, normalization, clustering, and a UMAP. Explain each step.✓ Copiado
    → Claude calls the right Scanpy functions in the right order with sensible defaults
  2. Iterate on parameters
    Re-do clustering with resolution 0.8 and show how cluster labels changed.✓ Copiado
    → Parameter tweak without re-planning the full pipeline

Resultado: A reproducible notebook plus figures, using idiomatic Scanpy — no hallucinated function names.

Errores comunes
  • Generic ML advice from Claude that ignores domain conventions — Reference the skill by name: 'use the scanpy skill'
Combinar con: filesystem

Build a structured literature review from PubMed + arXiv

👤 PhD students, research scientists ⏱ ~45 min intermediate

Cuándo usarlo: You need 30+ papers on a topic with abstracts and citation metadata, not just titles from Google.

Flujo
  1. Specify scope
    Use the paper-lookup skill. Find all PubMed and bioRxiv papers on GLP-1 agonists for Alzheimer's in the last 3 years.✓ Copiado
    → Claude hits the right APIs with correct query syntax
  2. Cluster by theme
    Cluster by hypothesis (neuroinflammation, vascular, direct neuronal) and give me the top 3 papers per cluster.✓ Copiado
    → Thematic grouping with citation-ready metadata

Resultado: A reviewable literature map with real DOIs, not fabricated citations.

Errores comunes
  • Claude hallucinates paper titles without the skill — Always verify DOIs — have Claude fetch one record to prove it's real
Combinar con: fetch

Dock a small molecule to a target protein with DiffDock

👤 Medicinal chemists, drug discovery researchers ⏱ ~20 min advanced

Cuándo usarlo: You have a SMILES string and a PDB target and want a quick first-pass pose prediction.

Requisitos previos
  • DiffDock environment — Follow the DiffDock skill's env-setup recipe
Flujo
  1. Provide ligand and receptor
    Using the diffdock skill, dock SMILES 'CC(=O)OC1=CC=CC=C1C(=O)O' into PDB 1ABC. Give me the top 5 poses with scores.✓ Copiado
    → Claude runs the right DiffDock command with correct flags
  2. Visualize
    Generate a PyMOL script to render the top pose.✓ Copiado
    → Runnable .pml file

Resultado: Pose predictions ready to feed into a downstream free-energy calculation.

Errores comunes
  • Claude invents flags that don't exist in DiffDock — The skill's reference folder has the real CLI — insist Claude consult it
Combinar con: filesystem

Combinaciones

Combínalo con otros MCPs para multiplicar por 10

scientific-agent-skill + filesystem

Skills produce notebooks and figures; filesystem MCP stores and organizes the output

Save the Scanpy UMAP to results/figures/ and write a README for the experiment.✓ Copiado
scientific-agent-skill + arxiv

Pair the paper-lookup skill with the arxiv MCP for richer citation graphs

Find arXiv papers citing the one I just read, grouped by which section they cite.✓ Copiado

Herramientas

Lo que expone este MCP

HerramientaEntradasCuándo llamarCoste
Bioinformatics (BioPython, Scanpy, pysam, gget, scVelo) sequence / count matrix / BAM file Sequence, genomics, or single-cell workflow 0 — local compute
Cheminformatics (RDKit, Datamol, DeepChem, DiffDock, OpenMM) SMILES / PDB / MOL2 Small molecule, protein structure, or dynamics problem 0
Clinical databases (ClinVar, COSMIC, ClinicalTrials.gov, FDA) gene / variant / trial ID Looking up clinical evidence or trial status 0 — public APIs
Paper lookup (PubMed, bioRxiv, arXiv) query string, date range Literature search with real citations 0
ML training (PyTorch Lightning, Transformers, PyMC, TimesFM) dataset + config Building a model with idiomatic framework usage 0

Coste y límites

Lo que cuesta ejecutarlo

Cuota de API
Public databases (NCBI, PubMed) have their own polite-use limits — typically 3 req/s
Tokens por llamada
SKILL.md references are sizeable; expect 2-5k tokens loaded per domain touched
Monetario
Free — skills are local files; only pay for compute you run
Consejo
Scope the prompt to one domain at a time so Claude doesn't load every SKILL.md.

Seguridad

Permisos, secretos, alcance

Almacenamiento de credenciales: No credentials in the skill itself; NCBI/PubMed work better with an email-in-env-var for polite rate limiting
Salida de datos: Only to the public science APIs you choose to query

Resolución de problemas

Errores comunes y soluciones

API rate limit hit on NCBI

Add an email to NCBI_EMAIL env var and cap concurrency at 3 req/s.

Skill not invoked on a relevant prompt

Mention the library name explicitly ('use the rdkit skill') — auto-invocation is fuzzy across 133 skills.

Alternativas

scientific-agent-skills vs otros

AlternativaCuándo usarlaContrapartida
ai-research-skill (Orchestra-Research)You want ML-research infrastructure (training, distributed, inference) rather than wet-lab / clinicalMore ML systems, less biology / chemistry domain depth
biomcpYou want an MCP server (live tools) rather than prompt-bundle skillsMCPs execute calls; skills just teach Claude to write the right code

Más

Recursos

📖 Lee el README oficial en GitHub

🐙 Ver issues abiertas

🔍 Ver todos los 400+ servidores MCP y Skills