healthtech-advisor
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided product descriptions via a Python script to identify HIPAA scope. This creates an ingestion point for untrusted data that is subsequently interpreted by the agent.
- Ingestion points: The script
scripts/phi_scope_checker.pyreads a user-provided text file (e.g.,description.txt) as part of its core workflow. - Boundary markers: The skill instructions do not define specific delimiters or warnings to prevent the agent from obeying instructions that might be embedded within the untrusted description text.
- Capability inventory: The
phi_scope_checker.pyscript performs read-only operations on the input file and outputs analysis to stdout; it contains no network, file-write, or dynamic execution capabilities, limiting the impact of successful injection. - Sanitization: No sanitization or escaping is performed on the input text before it is processed by the regex engine or reflected in the tool's output to the agent.
Audit Metadata