create-verification-skill

Pass

Audited by Gen Agent Trust Hub on Nov 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed to automate project verification by generating boilerplate for launching, driving, and capturing evidence from an application. It relies on standard system tools (bash, grep, glob, view, create, edit) to inspect the codebase and generate documentation/scripts.
  • [SAFE]: The 'Interview' and 'Prove' phases follow local development best practices, such as using isolated data directories and ports to prevent interference with other user sessions.
  • [SAFE]: No obfuscation, data exfiltration, or unauthorized privilege escalation patterns were found. The skill operates strictly within the project scope (.agents/skills/verify-<app>/).
  • [SAFE]: References to external harnesses (Playwright, Cypress, etc.) are treated as dependencies for the user to provide locally, rather than remote scripts fetched and executed silently by the skill itself.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 11, 2026, 06:43 AM