spec-reviewer

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists entirely of instructional prompts without any associated executable scripts or configurations. It defines a structured methodology for identifying user flows and requirements gaps.
  • [SAFE]: The permitted tools (Read, Grep, Glob) are restricted to file discovery and reading, which are appropriate for a specification review use case and do not present an elevated security risk in this context.
  • [SAFE]: No indicators of prompt injection, data exfiltration, or persistence mechanisms were found. The skill follows best practices for secret management by not including any hardcoded credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 12:28 AM