people-search
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted person profile data from both user-supplied CSV files and external third-party APIs, which presents a surface for indirect prompt injection attacks.\n
- Ingestion points: Data is ingested through the
read_csvandprovider_requestfunctions inscripts/people_search.py.\n - Boundary markers: The skill does not employ explicit delimiters or instructions to ignore potential commands embedded within the profile data it processes.\n
- Capability inventory: The skill possesses network communication capabilities (
urllib.request) and local file-reading functionality.\n - Sanitization: Input data undergoes basic normalization and string conversion in
coerce_scalar, but the skill lacks specific sanitization or filtering logic for prompt injection patterns.\n- [COMMAND_EXECUTION]: The skill's test suite (tests/test_people_search.py) usessubprocess.runto invoke thescripts/people_search.pyscript for automated verification. This usage is a benign development practice and is restricted to the local execution of the skill's own scripts.
Audit Metadata