seam-craft
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a clear operational boundary that prioritizes local deterministic tools over external services, reducing the potential attack surface.
- [SAFE]: A robust safety protocol is established for external tool usage (
sandbase_run), mandating that the agent discloses all parameters, endpoints, and estimated costs to the user for approval before execution. - [SAFE]: The instructions emphasize data minimization, explicitly directing the agent to send only the minimum necessary data when interacting with external APIs.
- [SAFE]: The skill uses vendor-specific tools (
sandbase_discover,sandbase_inspect, etc.) that are consistent with the author's identity and the skill's stated purpose.
Audit Metadata