measure-instrumentation-spec
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and serves to guide the agent in generating a markdown-based instrumentation spec. It does not perform any network operations, file system modifications, or command execution.
- [SAFE]: The instructions include comprehensive privacy and security requirements, specifically mandating that users specify data classes, minimization techniques (at both egress and storage boundaries), and fail-closed behaviors for AI model trace capture.
- [SAFE]: Analysis of the skill body, templates, and history reveals no malicious patterns, obfuscation, or unauthorized credential access. The provided examples and history suggest a high degree of security awareness and hardening against privacy leaks.
Audit Metadata