scientific-cellxgene-census

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection within the code used to query the genomics database.\n
  • Ingestion points: The tissue, cell_type, and gene_symbols parameters in the census_cell_metadata, census_gene_expression, and celltype_distribution functions in SKILL.md.\n
  • Boundary markers: No delimiters or instructions are used to separate the user-provided filter values from the SOMA query logic.\n
  • Capability inventory: The skill is capable of performing network reads from the Census API and writing data to local CSV and H5AD files.\n
  • Sanitization: User inputs are directly interpolated into f-strings to build query filters, allowing for potential manipulation of the data request logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 10:48 PM
Security Audit — agent-trust-hub — scientific-cellxgene-census