bio-differential-expression-de-results
Pass
Audited by Gen Agent Trust Hub on Aug 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill performs standard bioinformatics data processing using established R libraries such as DESeq2, edgeR, and dplyr. The provided code examples simulate data locally and perform routine filtering, annotation, and file export tasks (CSV, Excel) without any suspicious behavior.
- [PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it ingests and processes external biological datasets. 1. Ingestion points: Data enters the agent via R data frames generated from DESeq2/edgeR objects or loaded from CSV/GFF annotation files. 2. Boundary markers: Data is constrained within structured R objects (data.frames), reducing the risk of control-flow manipulation. 3. Capability inventory: Capabilities are limited to local file system writes for results export and standard package management via BiocManager. 4. Sanitization: Standard data manipulation functions (filter, join) are used, which do not execute data as code.
Audit Metadata