We Counted Every Science Skill in the OpenClaw Ecosystem
The OpenClaw science skill ecosystem has exploded. But how deep is it really? Which research domains have mature tooling, and which are still waiting for their first useful skill?
We surveyed 12 science-focused skill hubs containing over 2,000 skills to find out. Here's what we found.
The Full Picture
| Domain | Total Skills | Status |
|---|---|---|
| Literature Search & Review | ~306 | Extremely Red Ocean |
| Data Analysis / Statistics / ML | ~237 | Deep Red Ocean |
| Clinical / Medical | ~216 | Deep Red Ocean |
| Genomics (WGS/WES/variant calling) | ~178 | Red Ocean |
| Academic Writing / LaTeX | ~118 | Red Ocean |
| Research Automation / Pipeline | ~109 | Red Ocean |
| Single-cell / scRNA-seq | ~102 | Red Ocean |
| Bulk RNA-seq / Transcriptomics | ~87 | Red Ocean |
| Drug Discovery / Cheminformatics | ~62 | Red Ocean |
| Proteomics | ~39 | Orange |
| Citation / Reference Management | ~34 | Orange |
| Structural Biology (protein/MD) | ~31 | Yellow |
| Research Ideation / Novelty Check | ~28 | Yellow |
| Spatial Transcriptomics | ~26 | Yellow |
| Metabolomics | ~26 | Yellow |
| Imaging / Microscopy | ~25 | Yellow |
| Ecology / Environmental | ~14 | Blue Ocean |
The Red Oceans: Don't Build Here Unless You're 10x Better
Literature Search (~306 skills)
Every single skill hub has literature search skills. PubMed alone has 8+ wrapper skills across different repos. arXiv has 6+. If you're building yet another "search PubMed" skill, you're competing with 300 existing ones.
What's missing though: Cross-database synthesis that actually works. Most skills search one database at a time. A skill that intelligently routes queries across PubMed, arXiv, bioRxiv, Semantic Scholar, and OpenAlex based on the research domain would be genuinely new.
Data Analysis / ML (~237 skills)
Commoditized. Every stats function, every visualization type, every ML framework has a skill wrapper. AIPOCH alone has 138 data analysis skills.
What's missing: Domain-specific interpretation. A skill that doesn't just run DESeq2 but interprets the results in the context of your specific biological question.
Clinical / Medical (~216 skills)
FreedomIntelligence (119 skills) and AIPOCH (55 skills) dominate. SOAP notes, clinical trials, adverse events, discharge summaries — all covered extensively.
The Yellow Zones: Room to Grow
Research Ideation / Novelty Check (~28 skills)
Only ARIS and K-Dense meaningfully address this. The gap: automated literature-gap detection. A skill that reads your draft and tells you "this specific claim has no supporting evidence in the last 5 years" would be transformative.
Spatial Transcriptomics (~26 skills)
One of the fastest-growing areas in genomics (Visium, MERFISH, Slide-seq), but skill coverage is thin relative to its research momentum. Most existing skills are basic tool wrappers. End-to-end spatial analysis pipelines are nearly absent.
Metabolomics (~26 skills)
Standalone metabolomics workflows are rare outside bioSkills. The few that exist are mostly database query wrappers, not analytical pipelines.
Structural Biology (~31 skills)
AlphaFold wrappers exist, but molecular dynamics simulation pipelines, free energy perturbation, and conformational analysis skills are sparse. As AI-driven drug design grows, this gap will become more painful.
Imaging / Microscopy (~25 skills)
Only K-Dense and bioSkills cover this area. Histopathology AI, whole-slide image analysis, and fluorescence microscopy quantification are barely represented.
The Blue Ocean: Wide Open
Ecology / Environmental (~14 skills)
The biggest gap in the entire ecosystem. Metagenomics, environmental DNA, population ecology, conservation genomics — all underserved. If you're an ecologist who can code, this is your chance to define the tooling for your entire field.
What This Means
For skill users: Literature search, data analysis, and clinical skills are well-served. You have plenty of choices. Use our curated picks to skip the evaluation work.
For skill developers: The low-hanging fruit is gone in literature search and data analysis. The real opportunities are in:
- Ecology / Environmental — nearly empty
- Imaging / Microscopy — barely started
- Spatial Transcriptomics — high demand, thin supply
- Research Ideation — automated gap detection is a killer feature waiting to be built
- Cross-domain synthesis — skills that work across multiple databases/tools simultaneously
For the ecosystem: The fact that 306 skills exist for literature search while only 14 exist for ecology tells us something about who's building skills (ML/bioinformatics researchers) vs who's not yet in the ecosystem (ecologists, environmental scientists, microscopists).
Methodology
- 12 skill hubs surveyed: K-Dense, FreedomIntelligence, Orchestra, Imbad0202, ClawBio, ARIS, STELLA, AIPOCH, AMiner, Research-Claw Plugins, OmicSkills, bioSkills
- Skills counted by listing repository contents and README category breakdowns
- Categories assigned based on skill names, descriptions, and parent directory names
- "Other" category (~200+ skills) includes physics, materials science, finance, and cross-domain tools
- Survey date: March 24, 2026
Data source: Claw4Science Skill Hubs — all 12 repositories are indexed and tracked.