product-scale-reference
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The core content of the skill consists of best practices and formatting advice for e-commerce imagery, focusing on reducing returns through accurate scale communication.
- [EXTERNAL_DOWNLOADS]: The 'Setup' section contains instructions to add a Model Context Protocol (MCP) server from
https://mcp.apimage.org/mcp. This domain is associated with the skill's author, 'apimageorg', and represents a standard extension mechanism for AI agents. - [CREDENTIALS_UNSAFE]: The setup command includes an
Authorizationheader with the placeholder valuesk_your_api_key. This is a standard documentation practice to indicate where a user should provide their own token and does not represent a credential leak. - [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for generating images that accepts external references.
- Ingestion points:
reference_imagesparameter in thegenerate_imagetool call (SKILL.md). - Boundary markers: None present in the prompt template.
- Capability inventory: Image generation via
flux-2-promodel. - Sanitization: None detected for the image reference inputs, though the risk is negligible as it pertains to visual synthesis rather than instruction parsing.
Audit Metadata