portfolio-story-architecture-by-13qd91c
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to read and process the contents of
references/cases.md, which contains detailed prompt excerpts and descriptions intended to influence the agent's output style. While this structure uses data to provide instructions, the risk is mitigated by explicit instructions in theSKILL.mdworkflow for the agent to distinguish observed evidence from its own recommendations and to use a 'Preserve/Replace/Avoid' contract to maintain boundaries. - [EXTERNAL_DOWNLOADS]: The
references/cases.mdfile contains numerous links to external media and case study documentation hosted ongoodcase.aiandx.com. The agent is instructed to use available tools to inspect these resources for visual verification. These references are part of the primary design assistance purpose and target well-known media hosting and social platforms.
Audit Metadata