snap-out-of-it
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-supplied problem descriptions, constraints, and proposed approaches to generate a critique.
- Ingestion points: The skill body in
SKILL.mdinstructs the leading agent to supply the problem and constraints provided by the user to the reviewer. - Boundary markers: There are no explicit delimiters or specific instructions to the reviewer to ignore potential instructions embedded within the user's problem description.
- Capability inventory: The skill invokes an external model family via a selection policy to perform the review.
- Sanitization: No sanitization or validation of the user-provided content is performed before passing it to the secondary agent.
Audit Metadata