grand-slam-offer
Pass
Audited by Gen Agent Trust Hub on Mar 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses web_search to gather product features, pricing, and competitive intelligence. This is a legitimate research function aligned with the skill's purpose.
- [PROMPT_INJECTION]: The skill processes untrusted external data from web searches and user-provided descriptions, creating a surface for potential indirect prompt injection. Ingestion points: Data retrieved via web_search and user-defined product fields. Boundary markers: None identified in the workflow to separate external data from instructions. Capability inventory: Limited to web_search; no file-writing or command execution capabilities detected. Sanitization: No specific validation or filtering of external content is mentioned.
Audit Metadata