steelman
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied input to define the argument target, which is then passed to a text-generation process. The skill provides explicit instructions to sanitize this input by refusing topics that are harmful, illegal, or discriminatory.
- Ingestion points: User's most recent message (target).
- Boundary markers: None explicitly defined in the prompt interpolation.
- Capability inventory: Text generation via subagent spawning; no file, network, or code execution capabilities.
- Sanitization: Includes a mandatory safety filter that instructs the agent to refuse harmful, illegal, or violent targets.
Audit Metadata