amazon-product-photography
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill configuration and instructions do not contain any malicious patterns, obfuscated text, or dangerous command execution vectors. All references point to verified vendor resources or official Amazon documentation.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a text-based workflow that ingests external data, presenting a minor risk surface for indirect prompt injection.\n
- Ingestion points: Processes external data including product photos, listing URLs, packaging files, brand guides, competitor pages, and customer feedback inside
SKILL.md.\n - Boundary markers: Absent; no explicit delimiters or boundary isolation instructions are provided for untrusted external inputs.\n
- Capability inventory: None; the skill operates entirely as a text generation prompt and does not execute system commands, modify files, or perform network operations at runtime.\n
- Sanitization: Absent; no filtering or validation rules are defined for external data sources.
Audit Metadata