research / skill
Scientific Agent Skills
A library of skills that give an AI coding agent documented workflows for science packages, databases, and research tasks.
At a glance
- Purpose
- A library of skills that give an AI coding agent documented workflows for science packages, databases, and research tasks.
- Runs in
- Local agent with shell
- Operating systems
- macOS, Linux, Windows
- Needs
- Runtime: uv; Needs internet: Yes; Needs accounts or API keys beyond your agent: Not stated
- Files and data
- Reads files: Yes; Writes files: Yes; Sends data to other services: Yes
- Tool type
- Skill
- Delivery
- From GitHub
- Version
- Latest on the main branch
- Access
- Code on GitHub
- Compatibility
- According to the README, the collection works with Cursor, Claude Code, Claude Cowork, Codex, Gemini CLI, Google Antigravity, GitHub Copilot, VS Code, Kiro, OpenClaw, NemoClaw, Pi, Hermes, and other hosts that support the open Agent Skills standard.
- External costs
- The repository is MIT-licensed and states no price for itself. According to the README, individual skills connect to named external databases and services, and its Database Lookup skill documents 80 databases with public, registered, or licensed access; the README does not say which of them need an account or a licence. Each skill lists its own credential and network requirements. The README also says commercial support is available, without prices.
- Permissions
- The skills act through your coding agent with the access you give it. According to the README, skills can run Python code through packages installed with uv and make network requests to the databases and APIs they document. The README warns that skills can instruct your agent to run arbitrary code, install packages, make network requests and modify files on your system, and says to review what you install. Its steps for installing uv on macOS and Linux pipe a downloaded install script into sh, and its install routes place skill files in your agent's skills folders, for example ~/.agents/skills/.
- Usage terms
- MIT License, as stated in the repository. See its LICENSE.md file for the exact terms.
Code on GitHub
K-Dense-AI/scientific-agent-skills
MITTurn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 250,000+ scientists worldwide. 177 ready-to-use validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigrav
Topics: agent-skills ai-scientist bioinformatics chemoinformatics claude claude-skills claudecode clinical-research
Public GitHub data fetched · 0 days old
Skill folders in this repository (177)
Folders that contain a SKILL.md file, from GitHub's file list on .
skills/13c-metabolic-fluxskills/adaptyvskills/aeonskills/alphagenomeskills/analytical-method-validationskills/anndataskills/arborskills/arboretoskills/astropyskills/autoskillskills/benchling-integrationskills/bgpt-paper-searchskills/bidsskills/biopythonskills/bioservicesskills/bulk-rnaseqskills/canteraskills/cellprofilerskills/cellxgene-censusskills/cirqskills/citation-managementskills/clinical-decision-supportskills/clinical-reportsskills/cobrapyskills/consciousness-councilskills/daskskills/database-lookupskills/dataladskills/datamolskills/deepchemskills/deepspot-mskills/deeptoolsskills/depmapskills/dhdna-profilerskills/diffdockskills/dnanexus-integrationskills/esmskills/etetoolkitskills/exa-searchskills/experimental-designskills/exploratory-data-analysisskills/fictivskills/flowioskills/flowkitskills/fluidsimskills/folklore-variant-evidenceskills/generate-imageskills/genimlskills/genomic-coordinatesskills/genomic-intelligenceskills/geomasterskills/geopandasskills/get-available-resourcesskills/ggetskills/ginkgo-cloud-labskills/glycoengineeringskills/gtarsskills/histolabskills/hugging-scienceskills/hypogenicskills/hypothesis-generationskills/imaging-data-commonsskills/infographicsskills/iso-standards-readinessskills/lab-hardware-cadskills/labarchive-integrationskills/lamindbskills/latchbio-integrationskills/latex-postersskills/liteparseskills/literature-reviewskills/mageckskills/marine-carbonate-chemistryskills/markdown-mermaid-writingskills/market-research-reportsskills/markitdownskills/matchmsskills/matlabskills/matplotlibskills/medchemskills/modalskills/molecular-dynamicsskills/molfeatskills/ncats-araxskills/networkxskills/neurokit2skills/neuropixels-analysisskills/nextflowskills/nmrglueskills/nwb-conversionskills/omero-integrationskills/onekgpdskills/ontology-term-resolutionskills/open-notebookskills/openpivskills/opentrons-integrationskills/optimize-for-gpuskills/pacsomaticskills/paper-lookupskills/paperclip
And 77 more in the repository.
What it does
Scientific Agent Skills is a collection of skills for AI coding agents, created by K-Dense. According to the README, it covers biology, chemistry, medicine, physics, engineering, Earth science, data analysis, and scientific communication, with 177 skills, each documenting the conventions and validation checks for one package, data source, or workflow.The README groups the skills by area: bioinformatics and genomics, cheminformatics and drug discovery, proteomics, clinical research, healthcare AI and biosignal research, microscopy and medical imaging, neuroscience, machine learning, materials science, physics and astronomy, engineering and simulation, data analysis and visualization, geospatial science, laboratory automation, scientific communication, multi-omics, protein engineering, and regulatory or standards documentation. A dedicated Database Lookup skill documents access to 80 named scientific and financial databases.
According to the README, the agent can already use any Python package or API on its own; these skills add documentation and examples for the ones listed. The README also says it runs weekly LLM-based security scans on the skills in the repository and recommends reading each SKILL.md and scanning a skill yourself before installing it, since community contributions are not all reviewed to the same depth.
Who it is for
Scientists, researchers, and engineers who want an AI coding agent to follow documented workflows for specific scientific packages, databases, or research tasks.Example tasks
Query a chemical database for candidate compounds and analyze structure-activity relationships.Load single-cell RNA-seq data and run quality control and cell-type analysis.
Draft a costed experimental protocol from a planned experiment.
Search the literature and generate evidence-bound hypotheses for a research question.
Build a publication-quality figure, slide, or poster from a dataset.