cellxgene-census

Query the CZ CELLxGENE Census programmatically for versioned public single-cell and spatial transcriptomics data. Use when you need population-scale cell metadata, gene expression slices, Census summary counts, source H5AD URIs/downloads, embeddings, spatial Census data, or reference atlas comparisons across organisms, tissues, diseases, assays, and cell types. For analyzing your own local single-cell data use scanpy, anndata, or scvi-tools.

Quality

83%

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.

A well-structured, highly actionable skill body with executable examples and clearly signaled one-level-deep references. Its main weakness is verbosity from redundancy between inline best-practices/use-cases and the referenced patterns file.

Suggestions

Consolidate the inline 'Key Concepts and Best Practices' and 'Common Use Cases' sections into references/core_workflow_patterns.md to remove redundancy with that file and shorten the body.

Trim over-explanatory field annotations (e.g., the Boolean gloss on is_primary_data) since Claude can infer them from context.

Add an explicit validation/checkpoint note to the PyTorch training use case (e.g., inspect a batch's labels and shapes before the training loop) since it is a batch operation.

Dimension	Reasoning	Score
Conciseness	The ~290-line body is mostly useful but redundant: inline 'Key Concepts and Best Practices' and 'Common Use Cases' overlap with the referenced core_workflow_patterns.md, and some field explanations (e.g., 'is_primary_data (Boolean: True = unique cell)') over-explain what Claude can infer.	3 / 5
Actionability	Provides copy-paste-ready executable code throughout — open_soma, get_obs, get_anndata, get_presence_matrix, and a full TileDB-SOMA-ML PyTorch dataloader — covering the common query and training cases.	5 / 5
Workflow Clarity	Sequences multi-step work explicitly ('Two-Step Workflow: Explore Then Query') with a size-estimation checkpoint before loading (>100k cells → out-of-core), though the batch ML training example lacks an explicit validation/feedback step.	4 / 5
Progressive Disclosure	Three real reference files are each signaled with a 'When to read' clause at one level of depth, but some inline content (full use cases and troubleshooting) could be offloaded to references to tighten the overview.	4 / 5
	Total	16 / 20 Passed

Description

92%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.

A strong, specific description that clearly states what the skill does and when to use it, with effective boundary guidance against overlap with local-data analysis tools. The only minor weakness is that the trigger enumeration is feature-list-like rather than conversational.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — querying for cell metadata, gene expression slices, summary counts, source H5AD URIs/downloads, embeddings, spatial data, and reference atlas comparisons — giving comprehensive coverage of the skill's capabilities.	5 / 5
Completeness	Explicitly answers both 'what' ('Query the CZ CELLxGENE Census programmatically for versioned public single-cell and spatial transcriptomics data') and 'when' ('Use when you need...') with concrete trigger phrases and boundary guidance.	5 / 5
Trigger Term Quality	Includes natural terms users would say (single-cell, spatial transcriptomics, cell metadata, gene expression, embeddings) with good synonym coverage, though the long data-type enumeration reads more as a feature list than everyday user phrasing.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (public CZ CELLxGENE Census data) with explicit boundary guidance directing local-data analysis to scanpy, anndata, or scvi-tools, minimizing conflict risk.	5 / 5
	Total	19 / 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: K-Dense-AI/scientific-agent-skills
Path: skills/cellxgene-census/SKILL.md
Commit: 9cf7d9a

Reviewed: 19 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.