shopping-hunter
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from arbitrary external websites and writes the output to a local file, creating a vulnerability surface for indirect prompt injection.\n- Ingestion points: Reference links provided by the user and web search results retrieved during the parallel research passes described in SKILL.md.\n- Boundary markers: Absent; the instructions lack explicit delimiters or warnings to disregard potential commands embedded within external product data.\n- Capability inventory: The skill utilizes file writing capabilities to create a "Markdown dossier" in the local session files folder.\n- Sanitization: Absent; the skill does not specify procedures for sanitizing or escaping retrieved web content before including it in the dossier.
Audit Metadata