Skills
Navigate
skills/aipoch/medical-research-skills/roc-diagnostic-performance/Gen Agent Trust Hub

roc-diagnostic-performance

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external CSV and TSV files via the --expression_file and --group_file arguments in scripts/io.R. It lacks explicit boundary markers or specific instructions to ignore embedded instructions within the processed data. The skill possesses capabilities to write output files (CSV, RDS) and generate PDF plots (scripts/io.R, scripts/plotting.R). Sanitization is implemented in scripts/functions.R through mandatory gene matching against matrix headers and verification that all expression values are finite numeric types.
  • [DYNAMIC_EXECUTION]: The skill dynamically constructs R model formulas from the user-provided --marker_genes argument. This formula is then parsed using stats::as.formula in scripts/functions.R to fit the logistic regression model. This represents a dynamic execution pattern based on user-supplied strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 04:44 PM