tooluniverse-single-cell

Single-cell RNA-seq analysis with scanpy/anndata — h5ad data loading, scRNA-seq quality control and QC gating (n_genes_by_counts, total_counts, mitochondrial percent / pct_counts_mt, pct_counts_ribo, doublet detection with Scrublet/scDblFinder, ambient RNA / SoupX awareness, empty-droplet filtering, MAD-based thresholds), normalization, dimensionality reduction (PCA, UMAP, t-SNE), clustering (Leiden, Louvain), marker gene identification, cell-type annotation, pseudotime/trajectory analysis. Use for any scRNA-seq workflow, including deciding which cells to filter, flag, or investigate before downstream analysis.

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

88%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a strong, actionable, well-sequenced skill with real bundle files and clear navigation, including validation feedback loops for the QC gating workflow. Its main weaknesses are some redundant explanatory prose Claude likely already knows and a dangling reference to a missing analysis_patterns.md file.

Suggestions

Create the referenced analysis_patterns.md (or fix the decision-tree links to point to existing references such as marker_identification.md / cell_communication.md) so every cited reference resolves to a real bundle file.

Trim the Reasoning Framework prose that re-explains widely-known concepts (pseudo-bulk vs single-cell DE rationale, standard Leiden resolution ranges) down to skill-specific guidance Claude would not already apply.

Consider moving the biological 'WHY each metric flags a problem' detail into references/scrna_qc.md and keeping only the actionable metric->concern->action table inline to reduce body length.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with executable code and decision trees, but some prose re-explains concepts Claude already knows (e.g., the pseudo-bulk vs single-cell DE rationale and standard cluster-resolution guidance in the Reasoning Framework), placing it between the score-3 and score-5 anchors — noticeably above the midpoint but with trimmable explanation.	4 / 5
Actionability	Fully executable, copy-paste-ready code blocks (data loading, QC metrics, full pipeline, statistical tests, batch correction) plus specific commands (python scripts/scrna_qc.py, tu run run_deseq2_analysis) and a Scanpy-vs-Seurat table, covering the common cases per the score-5 anchor.	5 / 5
Workflow Clarity	Clear sequenced pipeline (QC -> Normalize -> HVG -> PCA -> Cluster -> Annotate -> DE) with explicit validation checkpoints and feedback loops (RULE ZERO check-first, HONEST EXECUTION install-plan/stop fallback, 'visualize distributions first', per-step removal reporting), matching the score-5 anchor including error-recovery guidance for the batch-like cell-filtering operation.	5 / 5
Progressive Disclosure	Well-organized overview with clearly signaled one-level-deep references to real files (references/scrna_qc.md, scanpy_workflow.md, etc.) and a dedicated Reference Documentation section, but analysis_patterns.md is referenced repeatedly in the decision tree and does not exist in the bundle — a dangling reference that keeps it just below the score-5 anchor.	4 / 5
	Total	18 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is exemplary: third-person, concrete, comprehensive in both capabilities and natural trigger terms, and explicitly pairs a 'what' list with a 'Use for...' 'when' clause. It is clearly distinct from sibling bulk/variant/enrichment skills.

Dimension	Reasoning	Score
Specificity	Lists many concrete actions (h5ad loading, QC gating with named metrics like n_genes_by_counts/pct_counts_mt, normalization, PCA/UMAP/t-SNE, Leiden/Louvain clustering, marker ID, annotation, pseudotime) — comprehensive coverage matching the score-5 anchor.	5 / 5
Completeness	Explicitly answers both 'what' (a long concrete action list) and 'when' via an explicit 'Use for any scRNA-seq workflow, including deciding which cells to filter, flag, or investigate before downstream analysis' clause with concrete triggers, matching the score-5 anchor.	5 / 5
Trigger Term Quality	Comprehensive natural terms including synonyms (single-cell RNA-seq / scRNA-seq), file extension (h5ad), and common user phrasings (clustering, marker genes, trajectory, QC), matching the score-5 anchor for synonym and extension coverage.	5 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (single-cell RNA-seq with scanpy/anndata) with distinct triggers and minimal conflict risk; the body even enumerates NOT-for boundaries against sibling skills, matching the score-5 anchor.	5 / 5
	Total	20 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: mims-harvard/ToolUniverse
Path: plugin/skills/tooluniverse-single-cell/SKILL.md
Commit: 1b02920

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

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.