pets-and-animals-lead
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the agent to browse and inspect media from external domains including
goodcase.ai,media.goodcase.ai, andx.com. These resources are used as primary evidence for visual prompt generation and storyboarding. - [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes external case summaries and prompt excerpts from
references/cases.md. This data acts as an attack surface where instructions embedded in external descriptions could influence the agent's behavior. The skill mitigates this by requiring a structured 'Reference Contract' and a verification phase to align the output with specific structural traits. - Ingestion points:
references/cases.mdcontains external summaries and prompt snippets from 30+ creators. - Boundary markers: The workflow mandates a 'Preserve / Replace / Avoid' contract to delimit which elements from the external source are utilized.
- Capability inventory: The skill utilizes image, video, and browser tools to inspect external media.
- Sanitization: Step 9 of the workflow requires manual comparison and revision to ensure output fidelity to the intended anchor case.
Audit Metadata