amazon-seasonal-planning
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No evidence of malicious behavior, obfuscation, or data exfiltration was found during the analysis. The skill's instructions align with its stated purpose of providing Amazon selling guidance.
- [NO_CODE]: The skill does not contain any scripts, binary files, or external code dependencies, consisting entirely of informational markdown documentation.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided product information and utilizes the
web_searchtool, which presents a surface for indirect prompt injection. This is evaluated as safe as the skill lacks high-privilege capabilities or access to sensitive local data. 1. Ingestion points: User input collection (SKILL.md); 2. Boundary markers: Absent; 3. Capability inventory: Use ofweb_searchfor event dates and rules; 4. Sanitization: Absent.
Audit Metadata