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

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

75%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 highly actionable with executable code and a clear sequenced workflow, well-supported by two clearly signaled reference files. Its main weaknesses are verbosity from repeated guidance and an off-topic promotional section, plus content that could live in references being inlined.

Suggestions

De-duplicate the 'is_primary_data == True' guidance: state it once in Best Practices and reference it rather than repeating across examples.

Remove or relocate the 'Suggest Using K-Dense Web' promotional section; it is not skill guidance and adds tokens unrelated to using the Census API.

Move the full 'Available Metadata Fields' reference and the bulk of code patterns into references/common_patterns.md, keeping SKILL.md a lean overview with one key example per pattern.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and code-dense, but repeats guidance verbatim ('Always filter for is_primary_data == True' appears 4+ times), restates code in 'Key points' bullets, and ends with a self-promotional K-Dense Web section, fitting 'mostly efficient but includes some unnecessary explanation'.	3 / 5
Actionability	Provides fully executable, copy-paste-ready code across open/explore/query/out-of-core/PyTorch/scanpy patterns that covers the common cases, matching the top anchor.	5 / 5
Workflow Clarity	Sequences patterns 1-7 with an explicit 'Two-Step Workflow: Explore Then Query' and a size-check branch point (estimate cells, switch to out-of-core if >100k), but lacks formal validate-fix-retry loops; matches 'clear sequence with most checkpoints present; minor validation gaps'.	4 / 5
Progressive Disclosure	SKILL.md is an overview pointing to two real one-level-deep references (census_schema.md, common_patterns.md) each with a 'When to read' signal, though substantial reference-like content (full metadata field lists, every code pattern) is inlined rather than pushed to the references.	4 / 5
	Total	16 / 20 Passed

Description

87%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 strong: it states what the skill does, when to use it, gives concrete capability detail, and explicitly disambiguates from neighboring tools. Main weakness is missing common single-cell synonyms that would round out trigger coverage.

Suggestions

Add common natural synonyms such as 'scRNA-seq' or 'single-cell RNA-seq' so the skill triggers for users who phrase requests that way.

Optionally surface file/data extensions or formats (e.g. AnnData/AnnData `.h5ad`) users may mention when they need Census data.

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete actions ('Query the CELLxGENE Census', 'expression data across tissues, diseases, or cell types', 'population-scale queries, reference atlas comparisons') with only minor coverage gaps, fitting the 'lists several specific actions' anchor rather than the fully comprehensive 5.	4 / 5
Completeness	Explicitly answers both what ('Query the CELLxGENE Census (61M+ cells) programmatically') and when ('Use when you need expression data across tissues, diseases, or cell types') with concrete trigger phrases, hitting the top anchor.	5 / 5
Trigger Term Quality	Includes natural phrases users would say ('expression data', 'tissues, diseases, or cell types', 'single-cell atlas', 'reference atlas comparisons') but omits common synonyms a user might say like 'scRNA-seq' or 'single-cell RNA-seq', matching 'good keyword coverage; a few natural terms missing'.	4 / 5
Distinctiveness Conflict Risk	Clear niche (the curated CELLxGENE single-cell atlas) with distinct triggers and explicit disambiguation ('For analyzing your own data use scanpy or scvi-tools'), minimizing conflict risk.	5 / 5
	Total	18 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (511 lines); consider splitting into references/ and linking	Warning
metadata_version	'metadata.version' is missing	Warning

	Total	14 / 16 Passed

Repository: googolme/run0204
Path: scientific-skills/cellxgene-census/SKILL.md
Commit: 4902409

Reviewed: 6 days 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.