NYC

offer-validator

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): The skill consists entirely of markdown instructions and metadata. No scripts, binaries, or network-enabled components are present.
  • [Indirect Prompt Injection] (LOW): The skill is designed to ingest and analyze user-provided business offer descriptions. This creates a surface for indirect prompt injection where a user might include instructions to bypass scoring criteria.
  • Ingestion points: User input via "Quick Gather Questions" or direct offer descriptions.
  • Boundary markers: Absent; the skill relies on natural language parsing.
  • Capability inventory: No subprocess calls, network access, or file-system modifications detected.
  • Sanitization: None; the AI is expected to process the input as-is for evaluation purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:31 PM