modernize-scientific-stack
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (LOW): The skill directs the agent to perform a Verification Protocol in Phase 4 that involves executing commands such as 'python -m py_compile', 'python -c', and functional script execution. This results in the execution of code generated from potentially untrusted inputs.
- [PROMPT_INJECTION] (LOW): The skill is susceptible to indirect prompt injection as it ingests untrusted legacy source code in Phase 1 without boundary markers or sanitization. Evidence Chain: 1. Ingestion points: Legacy source files (Phase 1). 2. Boundary markers: Absent. 3. Capability inventory: Python shell execution for syntax and functional verification (Phase 4). 4. Sanitization: None.
- [EXTERNAL_DOWNLOADS] (SAFE): The skill suggests migrating to industry-standard libraries such as 'pandas' and 'numpy' from trusted repositories.
Audit Metadata