cellxgene-census

Query the CELLxGENE Census (61M+ cells) programmatically. Use when you need expression data across tissues, diseases, or cell types from the largest curated single-cell atlas. Best for population-scale queries, reference atlas comparisons. For analyzing your own data use scanpy or scvi-tools.

Quality

70%

Does it follow best practices?

Impact

—

No eval scenarios have been run

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./scientific-skills/cellxgene-census/SKILL.md

Quality

Discovery

89%

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 a strong skill description that clearly identifies a specific niche (CELLxGENE Census querying), provides good trigger terms for the bioinformatics domain, and explicitly states both when to use it and when not to use it. The main weakness is that the specific actions/capabilities could be more granular — listing concrete operations like 'retrieve expression matrices, filter by metadata, download AnnData objects' would strengthen it further.

Suggestions

Add 2-3 more specific concrete actions beyond 'query', such as 'retrieve expression matrices, filter cells by metadata, download AnnData/sparse arrays' to improve specificity.

Dimension	Reasoning	Score
Specificity	Names the domain (CELLxGENE Census, single-cell data) and a general action ('Query...programmatically'), but doesn't list multiple specific concrete actions like 'retrieve gene expression matrices, filter by cell metadata, compute differential expression'. The mention of 'expression data across tissues, diseases, or cell types' adds some specificity but remains at a high level.	2 / 3
Completeness	Clearly answers 'what' (query CELLxGENE Census for expression data across tissues/diseases/cell types) and 'when' ('Use when you need expression data across tissues, diseases, or cell types', 'Best for population-scale queries, reference atlas comparisons'). Also includes a negative trigger ('For analyzing your own data use scanpy or scvi-tools') which helps Claude know when NOT to use it.	3 / 3
Trigger Term Quality	Includes strong natural keywords users would say: 'CELLxGENE', 'Census', 'expression data', 'tissues', 'diseases', 'cell types', 'single-cell atlas', 'population-scale queries', 'reference atlas'. Also differentiates from related tools (scanpy, scvi-tools) which helps with routing. These are terms a bioinformatician would naturally use.	3 / 3
Distinctiveness Conflict Risk	Highly distinctive — CELLxGENE Census is a very specific resource, and the description explicitly differentiates from scanpy/scvi-tools for own-data analysis. The '61M+ cells' detail and 'largest curated single-cell atlas' further narrow the niche. Unlikely to conflict with other bioinformatics skills.	3 / 3
	Total	11 / 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 excellent, actionable code examples covering a wide range of CELLxGENE Census use cases, but is severely bloated with redundant content, unnecessary explanations, and repeated patterns. The same information (e.g., is_primary_data filtering, query patterns) appears in multiple sections. The document would benefit greatly from being cut to roughly one-third its current size, moving detailed patterns to the referenced files, and keeping only a concise overview with the most essential examples in SKILL.md.

Suggestions

Cut the 'Overview', 'When to Use This Skill', and 'Available Metadata Fields' sections entirely — the description already covers when to use it, and metadata fields belong in the schema reference file.

Consolidate 'Common Use Cases' into the 'Core Workflow Patterns' section — they are largely duplicative. Keep only 3-4 essential patterns in SKILL.md and move the rest to references/common_patterns.md.

Remove repeated best-practice advice from inline comments and the dedicated Best Practices section — state each tip once, ideally as a brief bullet list rather than full code blocks.

Add an explicit validation gate in the core workflow: after estimating query size, include a clear decision point (e.g., '>100k cells → use out-of-core; otherwise → get_anndata') as a numbered step rather than a suggestion buried in best practices.

Dimension	Reasoning	Score
Conciseness	Extremely verbose at ~350+ lines. Extensively repeats patterns (e.g., is_primary_data == True appears 15+ times), explains concepts Claude already knows (what Census includes, when to use the skill, basic scanpy workflows), and duplicates content between 'Core Workflow Patterns', 'Best Practices', and 'Common Use Cases' sections. The overview bullets, 'When to Use' section, and much of the troubleshooting are unnecessary padding.	1 / 3
Actionability	Provides fully executable, copy-paste ready code examples throughout. Filter syntax is clearly documented with concrete examples, API calls include all required parameters, and both small-scale (get_anndata) and large-scale (axis_query) patterns are complete and runnable.	3 / 3
Workflow Clarity	The two-step 'explore then query' workflow and the size-estimation checkpoint are good, but the overall document lacks a clear sequential workflow with explicit validation steps. The progression from opening Census → exploring → querying → processing is implicit rather than explicitly sequenced with checkpoints. Memory estimation is mentioned as a best practice but not integrated into the core workflow as a mandatory validation gate.	2 / 3
Progressive Disclosure	References to references/census_schema.md and references/common_patterns.md are well-signaled with clear 'when to read' guidance, but no bundle files were provided to verify these exist. The main SKILL.md itself is monolithic — much of the content in sections 5-7 and the Common Use Cases could be in reference files, and the Best Practices section heavily overlaps with inline guidance already given in the workflow patterns.	2 / 3
	Total	8 / 12 Passed

Validation

81%

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 — 9 / 11 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (510 lines); consider splitting into references/ and linking	Warning
metadata_version	'metadata.version' is missing	Warning

	Total	9 / 11 Passed

Repository: K-Dense-AI/claude-scientific-skills
Commit: cbcae7b

Reviewed: about 11 hours 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.