Official Bioclaw Skills Library for Bioinformatics and Omics Workflows
Bioclaw_Skills_Hub is a public collection of reusable skills for bioinformatics, omics analysis, and computational biology.
It is organized around real analysis tasks rather than isolated tools, making it easier to browse, reuse, and adapt for AI agents, workflow systems, and research assistants.
- Overview
- What You Will Find Here
- Repository Structure
- Major Domains
- What Makes This Repository Useful
- Project Status
- How To Use It
- Testing Skill Changes
- Design Principles
- License
Bioinformatics workflows are often scattered across many small prompt collections, tool-specific notes, and one-off agent instructions.
Bioclaw_Skills_Hub brings those patterns together into a more consistent repository structure. The goal is to provide a skill library that is easier to curate, easier to route, and easier to reuse across different omics settings, including transcriptomics, single-cell analysis, epigenomics, metagenomics, proteomics, structural biology, and general bioinformatics.
Complementary external resource: Paperzilla (ClawHub) is a useful companion when omics work depends on conversational literature monitoring. It currently supports project-based bioRxiv and medRxiv feeds, can fetch paper markdown for summarization and relevance assessment, and has PubMed support planned.
- task-focused skills for common omics workflows
- a taxonomy that groups skills by major analysis domain
- deeper references for high-value workflows such as ChIP-seq, ATAC-seq, differential expression, metagenomics, proteomics, structural biology, and single-cell analysis
- a foundation for building smaller runtime packs for agent systems such as BioClaw
Bioclaw_Skills_Hub/
├── skills/
│ ├── transcriptomics/
│ ├── single-cell-and-spatial/
│ ├── epigenomics-and-regulation/
│ ├── genomics-and-variation/
│ ├── metagenomics-and-microbiome/
│ ├── proteomics-and-metabolomics/
│ ├── multi-omics-and-systems/
│ ├── core-bioinformatics/
│ └── .../
├── catalog/
├── scripts/
└── .github/workflows/
- Transcriptomics
- Single-cell and spatial
- Epigenomics and regulation
- Genomics and variation
- Metagenomics and microbiome
- Proteomics and metabolomics
- Multi-omics and systems biology
- Core bioinformatics
- It reduces duplication across many small skill collections by consolidating them into a clearer omics taxonomy.
- It keeps high-level categories compact while still allowing specialized leaf skills.
- It is suitable both as a standalone public skill library and as a source repository for downstream agent packs.
This repository will continue to be iterated and expanded over time.
If you find it useful, consider giving the project a star ⭐.
This repository is inspired by earlier skill collections including claude-scientific-skills, OpenClaw-Medical-Skills, claude-for-life-sciences-skills and bioSkills.
The current project reorganizes and refines those ideas into a more unified omics-oriented library structure.
- Browse
skills/when you want a workflow-oriented entry point. - Use
catalog/when you want the compact taxonomy and source mapping. - Reuse individual leaf skills directly, or curate a smaller subset for a production agent environment.
Before pushing changes to skills/, run the local test pipeline:
python _bioclaw_test/run_tests.pyOnly push skill updates after this test passes. It checks taxonomy consistency, markdown links, absolute filesystem paths, and required reference files.
- Group skills by user-facing analysis goals, not just package names.
- Keep leaf skills actionable and easy to route.
- Separate broad workflow guidance from deeper technical references.
- Make it easy to derive a smaller, high-signal runtime pack from a larger public library.
This project is released under the MIT License.