wtf
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill defines a specific command-based interaction model for providing explanations and rationales. It contains no code, network operations, or file system access.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the conversation history. The risk of the agent following instructions embedded in that history is mitigated by the instruction to only clarify and 'do not implement, take side effects, or make new decisions.'
- Ingestion points: Conversation history and user-supplied topics referenced in SKILL.md
- Boundary markers: None specified
- Capability inventory: None; the skill explicitly restricts the agent from performing actions or side effects across all instructions
- Sanitization: Not specified
Audit Metadata