tooluniverse-crispr-screen-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a comprehensive Python-based implementation for functional genomics, including library normalization, quality control, and robust rank aggregation (MAGeCK-like) scoring. All code snippets in
ANALYSIS_DETAILS.mdandreference_gene_sets.pyuse standard libraries (pandas, numpy, scipy) for legitimate analytical purposes. - [SAFE]: The skill references established scientific and clinical data sources. These include PubMed (NCBI), Reactome, Enrichr, DGIdb, and Pharos. These are well-known, trusted services in the bioinformatics community.
- [SAFE]: The fallback mechanism detailed in
FALLBACK_PATCH.mdis a well-engineered solution for handling API downtime. It transparently shifts validation and essentiality proxies to the Open Targets Platform, illustrating a defensive and pragmatic design for a research tool. - [INDIRECT_PROMPT_INJECTION]: The skill processes external CRISPR count matrices (TSV/CSV) provided by the user. While this ingestion point theoretically allows for indirect prompt injection if malicious content is placed in the data files, the skill handles data through structured parsing (pandas) and standard scientific workflows. The risk is considered safe as it is inherent to any data-processing utility and no specific malicious exploitation patterns were found. Evidence chain: 1. Ingestion: Phase 1 (Count matrices). 2. Boundaries: Structured TSV/CSV format. 3. Capabilities: Report generation and API queries. 4. Sanitization: Standard data frame processing.
Audit Metadata