claude-scientific-skills

I took a look at your claude-scientific-skills skill and wanted to share some thoughts.

Links:

• GitHub
• Your skill in the SkillzWave marketplace

The TL;DR

You're at 50/100, which lands you in failing territory. The grading is based on Anthropic's Progressive Disclosure Architecture framework – how efficiently skills guide Claude through use. Your strongest area is Utility (15/20) – the individual sub-skills like RDKit and Scanpy are legitimately excellent. The biggest drag is Progressive Disclosure Architecture (12/30) – the way this is structured breaks some fundamental expectations about how skills should work.

What's Working Well

• Sub-skill quality is chef's kiss. Your RDKit (764 lines), Scanpy, and ChEMBL implementations are domain-expert quality with real code examples, error handling, and practical workflows. This is the heavy lifting that makes the collection genuinely useful.
• Clear terminology consistency. Within each domain (cheminformatics, bioinformatics, etc.), you use consistent language and don't mix paradigms.
• Comprehensive coverage. 127 individual scientific skills across genomics, drug discovery, proteomics – you're addressing real capability gaps that researchers actually need.

The Big One: This Isn't a Skill, It's a Skill Collection – And The Architecture Reflects That

Here's the core issue: Your SKILL.md at the top level is 59 lines of marketing copy pointing to a subdirectory with 127 other skills. That's not a Progressive Disclosure Architecture – that's an index. And it costs you 8 points because:

1. Claude can't invoke this meaningfully. When someone asks "help me do sequence alignment," Claude sees a generic index with no trigger phrases. Should it load scanpy/SKILL.md? biopython/SKILL.md? The skill doesn't guide that decision.

2. References go 2+ levels deep. SKILL.md → scientific-skills/rdkit/SKILL.md → references/. That violates the one-level reference depth principle – it breaks token economy by forcing Claude to traverse too many files.

3. No actionable workflow in the top-level. Your "Getting Started" is literally "explore the subdirectory" – there's no entrypoint for Claude.

The Fix: You have two paths:

• Option A (Recommended): Restructure as a skill registry. Make the top-level SKILL.md a proper index with a "Quick Navigation" section listing every sub-skill with 1-2 word descriptions and their triggers. Add a workflow like: "1. Identify your task (sequence analysis, molecular docking, etc.) 2. Load the corresponding skill from scientific-skills/ 3. Follow that skill's workflow."

• Option B: Accept this is a plugin/collection, not a skill. Submit each sub-skill individually to SkillzWave. You'll get 127 separate entries, which is more discoverable and each gets its own score.

Other Things Worth Fixing

1. Marketing language breaks spec compliance (–4 points). Your README and top SKILL.md use "Transform Claude into an AI Scientist," promotional badges, star history. Skill documentation should be instructional, not marketing. Replace with: "Collection of 127 scientific domain skills providing APIs for..." Keep the README as-is for GitHub visibility, but clean up SKILL.md.

2. Frontmatter needs trigger phrases (+2 points). Current description: "Comprehensive collection of 128+ ready-to-use scientific skills..." Add activation triggers: "Use when asked about RDKit, Scanpy, ChEMBL, molecular docking, single-cell RNA-seq, drug discovery, genomics analysis."

3. No examples or templates in top-level skill (+2 points). Your sub-skills have great code snippets. The main SKILL.md has zero. Add a quick "Common Tasks" section with 2–3 tiny examples showing how to invoke sub-skills.

Quick Wins

• Restructure top-level as a proper index with sub-skill navigation (biggest bang for buck – +8 points)
• Replace marketing copy with technical language (+3 points)
• Add trigger phrases to frontmatter description (+2 points)
• Include 2–3 quick examples showing skill invocation (+2 points)

Hit these four and you're pushing 65–70/100 easily.

Checkout your skill here: SkillzWave.ai | SpillWave We have an agentic skill installer that install skills in 14+ coding agent platforms. Check out this guide on how to improve your agentic skills.