chef-pro
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the user which could contain hidden instructions designed to influence the agent's behavior.
- Ingestion points: User-provided inputs during 'Step 1: รวบรวม restaurant profile' in SKILL.md (specifically Concept and Problem descriptions).
- Boundary markers: Absent. The instructions do not define delimiters or specific warnings to ignore instructions embedded within the restaurant profile data.
- Capability inventory: None detected. The skill generates text-based recommendations and does not utilize subprocess execution, file writes, or network requests.
- Sanitization: Absent. There are no explicit steps for the agent to validate or sanitize the incoming text data.
Audit Metadata