pitch
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill acts as a router for product marketing tasks, delegating work to specialized sub-skills like pitch-copy and pitch-recon. The implementation is consistent with its stated purpose and the tool usage is appropriately scoped in the metadata.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it passes untrusted user input directly to other tools. (1) Ingestion points: User input enters via the
{{args}}parameter inSKILL.md. (2) Boundary markers: The input is interpolated into instructions without delimiters or boundary markers. (3) Capability inventory: The skill metadata allows access to high-capability tools such asBash,Write, andWebFetch. (4) Sanitization: No input validation or filtering is applied to user arguments before processing.
Audit Metadata