product-photo-consistency
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection vulnerability surface as it is designed to process untrusted external data (product catalogues) and images (supplied photos).
- Ingestion points: Processes product metadata from
catalogueobjects and analyzes user-provided images viaanalyze_image. - Boundary markers: Absent; the prompt templates provided in the documentation do not include explicit delimiters or instructions to ignore embedded commands within the product data.
- Capability inventory: Accesses tools for image generation (
generate_image), image analysis (analyze_image), and image editing (remove_background,edit_image). - Sanitization: No sanitization or validation of the product metadata or image content is mentioned before interpolation into prompts.
- [COMMAND_EXECUTION]: The skill provides instructions for the user to execute a setup command to add an external MCP server (
claude mcp add). This command targets the author's own infrastructure (mcp.apimage.org) and uses a standard placeholder for authentication, which is consistent with the intended functionality of the vendor's skill.
Audit Metadata