payroll-compensation

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [NO_CODE]: The skill contains no executable code or scripts, consisting only of Markdown and YAML documentation for payroll processes.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its ingestion of external data fields in employee and contractor records.
  • Ingestion points: employee_record and contractor_payment data structures in SKILL.md.
  • Boundary markers: Lacks specific delimiters or instructions to ignore commands within the data fields.
  • Capability inventory: Orchestrates the salary-management sub-skill and performs calculations.
  • Sanitization: No input validation or sanitization is implemented for the data fields.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 07:50 PM