suede-free-tools
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of prompt instructions and reference documentation for business strategy. It does not perform network operations, file system modifications (beyond reading context), or command execution.
- [INDIRECT_PROMPT_INJECTION]: The skill reads from local context files such as
.agents/product-marketing.md. However, the instructions are scoped to generating structured textual recommendations, and it lacks dangerous capabilities like code execution or network exfiltration, minimizing the risk of exploitation. - [SAFE]: The skill includes a 'Boundaries' section that explicitly instructs the agent not to deploy code, publish content, or connect production data without user authorization, which serves as a safety guardrail.
Audit Metadata