cellxgene-census
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [Overall Security] (SAFE): The provided files are purely educational and reference-based documentation for the cellxgene_census Python library.
- [Data Exposure] (SAFE): No credentials or sensitive local paths are exposed. The library accesses public biological data from the CZ CELLxGENE repository.
- [Prompt Injection] (SAFE): No instructions targeting agent behavior override or safety filters were found.
- [Dependency Analysis] (SAFE): References standard, reputable scientific libraries including cellxgene-census, scanpy, torch, and pyarrow.
- [Indirect Prompt Injection] (INFO): The skill documents access to external scientific data. 1. Ingestion points: cellxgene_census.get_obs, get_anndata, and axis_query in common_patterns.md. 2. Boundary markers: Absent. 3. Capability inventory: Data processing and visualization using scientific libraries. 4. Sanitization: Absent. The risk is negligible as the data source is a curated genomic database and the processed data is numeric or standardized biological metadata.
Audit Metadata