scanpy
Standard single-cell RNA-seq analysis pipeline. Use for QC, normalization, dimensionality reduction (PCA/UMAP/t-SNE), clustering, differential expression, and visualization. Best for exploratory scRNA-seq analysis with established workflows. For deep learning models use scvi-tools; for data format questions use anndata.
79
75%
Does it follow best practices?
Impact
82%
1.51xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./scientific-skills/scanpy/SKILL.mdQuality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that clearly defines its scope within single-cell RNA-seq analysis, lists specific concrete actions, uses domain-appropriate trigger terms that users would naturally employ, and explicitly delineates boundaries with related skills. The inclusion of negative triggers ('For deep learning models use scvi-tools; for data format questions use anndata') is a particularly strong feature that minimizes conflict risk.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: QC, normalization, dimensionality reduction (PCA/UMAP/t-SNE), clustering, differential expression, and visualization. These are well-defined bioinformatics operations. | 3 / 3 |
Completeness | Clearly answers both 'what' (QC, normalization, dimensionality reduction, clustering, DE, visualization) and 'when' ('Use for... exploratory scRNA-seq analysis with established workflows'). Also provides explicit negative triggers directing to other skills (scvi-tools for deep learning, anndata for data format). | 3 / 3 |
Trigger Term Quality | Includes strong natural keywords users would say: 'scRNA-seq', 'QC', 'normalization', 'PCA', 'UMAP', 't-SNE', 'clustering', 'differential expression', 'single-cell RNA-seq'. These are the exact terms a bioinformatician would use. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche in single-cell RNA-seq analysis. Explicitly differentiates itself from related skills (scvi-tools for deep learning, anndata for data formats), which reduces conflict risk significantly. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
50%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill provides a comprehensive and actionable scanpy workflow with excellent executable code examples covering the full analysis pipeline. However, it is significantly too verbose—it explains concepts Claude already knows (AnnData structure, what scanpy is), duplicates content that exists in referenced files, and includes redundant advisory sections. The workflow would benefit from explicit validation checkpoints and tighter content that delegates details to the referenced materials.
Suggestions
Remove the 'When to Use This Skill' section (this belongs in frontmatter description), the AnnData structure explanation (Claude knows this), and the 'Additional Resources' links section to reduce token usage significantly.
Consolidate 'Common Pitfalls and Best Practices' and 'Tips for Effective Analysis' into a single concise section—they currently overlap substantially.
Add explicit validation checkpoints in the workflow, e.g., 'Verify filtered cell count is reasonable (expect 50-90% retention)' after QC, and 'If UMAP shows no structure, revisit n_pcs and n_neighbors' after dimensionality reduction.
Move the detailed standard workflow code to references/standard_workflow.md and keep only a minimal quick-start example in SKILL.md, since the reference file already contains the complete workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is excessively verbose. It explains AnnData structure (which Claude knows), includes a 'When to Use This Skill' section that restates the description, explains what scanpy is, and has redundant sections like 'Tips for Effective Analysis' that repeat 'Common Pitfalls and Best Practices'. The 'Understanding AnnData Structure' block and parameter tables explain concepts Claude already knows. Much content duplicates what's in the referenced files. | 1 / 3 |
Actionability | The skill provides fully executable, copy-paste ready code throughout. Every step of the workflow includes concrete Python code with specific function calls, parameters, and realistic examples. CLI commands for scripts are also complete with flags. | 3 / 3 |
Workflow Clarity | The workflow is clearly sequenced with numbered steps (1-7), but lacks explicit validation checkpoints. There are no feedback loops for error recovery—e.g., no guidance on what to do if QC filtering removes too many cells, or if clustering results look wrong. The 'Check QC plots carefully' advice is vague rather than being an explicit validation step in the workflow. | 2 / 3 |
Progressive Disclosure | References to external files are present and well-signaled (scripts/qc_analysis.py, references/standard_workflow.md, references/api_reference.md, references/plotting_guide.md, assets/analysis_template.py). However, the main SKILL.md contains too much inline content that overlaps significantly with the referenced files—the entire standard workflow is duplicated here and in references/standard_workflow.md, and plotting examples duplicate the plotting guide. | 2 / 3 |
Total | 8 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
metadata_version | 'metadata.version' is missing | Warning |
Total | 10 / 11 Passed | |
- Repository
- K-Dense-AI/claude-scientific-skills
- Commit
25e1c0f
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.