chro-advisor
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process data from external files, which creates a surface for potential instruction injection if the data contains malicious prompts.\n
- Ingestion points: The skill is instructed to read
company-context.md(if present) and the Python scripts (comp_benchmarker.pyandhiring_plan_modeler.py) ingest user-supplied JSON or CSV files containing employee rosters and hiring plans.\n - Boundary markers: The instructions do not implement specific delimiters or "ignore embedded instructions" warnings to handle potentially untrusted content within the company context or data files.\n
- Capability inventory: All included Python scripts were audited and found to be restricted to data processing, reporting, and standard console/CSV output. The skill lacks capabilities for network access, subprocess execution, or writing to the local filesystem.\n
- Sanitization: Input data is parsed using standard
jsonandcsvmodules without additional sanitization or escaping of string content before it is processed or presented to the agent.
Audit Metadata