social-carousel
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied briefs and executes instructions returned from the 'prepare_open_design_brief' tool, creating a surface for malicious content to influence agent actions.
- Ingestion points: User-provided messages (the brief) and tool output from 'prepare_open_design_brief' (SKILL.md).
- Boundary markers: None identified; the skill lacks delimiters or instructions to ignore malicious directives within the external content.
- Capability inventory: The agent is explicitly instructed to use its own file-editing tools to author files based on these dynamic instructions (SKILL.md).
- Sanitization: The skill does not specify any validation, filtering, or escaping of the instructions received from the tool before they are acted upon.
Audit Metadata