tooluniverse-single-cell
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides documentation for installing standard, well-known scientific Python packages from the official PyPI registry (scanpy, anndata, leidenalg, umap-learn, harmonypy, gseapy, pandas, numpy, scipy, scikit-learn, statsmodels).
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data files (h5ad, 10X, CSV, TSV) which represent a theoretical surface for indirect prompt injection if the files contain malicious metadata strings.
- Ingestion points: Data is loaded through functions like
sc.read_h5ad,sc.read_10x_h5, andpd.read_csvin the analysis scripts. - Boundary markers: The skill does not explicitly define boundary markers in its code snippets to separate data from instructions.
- Capability inventory: The skill utilizes the ToolUniverse framework for specialized genomic tasks and local Python execution for data analysis.
- Sanitization: There is no evidence of explicit sanitization for strings extracted from the genomic datasets before they are used in reporting.
Audit Metadata