clinical-decision-support
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a comprehensive suite of templates and scripts for clinical cohort analysis, treatment recommendations, and decision algorithms. The logic is consistent with the stated purpose.
- [EXTERNAL_DOWNLOADS]: The README.md recommends standard scientific dependencies (pandas, numpy, scipy, lifelines, matplotlib, pyyaml). These are well-known, trusted libraries for data science and are appropriate for this skill's functionality.
- [INDIRECT_PROMPT_INJECTION]: The skill scripts (biomarker_classifier.py, generate_survival_analysis.py) ingest external data in CSV and JSON formats. This is an expected attack surface for clinical analysis tools; however, the scripts perform legitimate data processing and statistical calculations without unsafe interpolation into agent instructions.
- [DYNAMIC_CONTEXT_INJECTION]: A review of SKILL.md confirms the absence of dynamic execution placeholders (!
command) at skill load time. - [COMMAND_EXECUTION]: The skill uses shell commands for legitimate purposes, such as running local analysis scripts and generating schematics. These operations are scoped to the project directory and do not exhibit malicious intent.
Audit Metadata