Skills
skills/wojons/skills/dogfooding/Gen Agent Trust Hub

dogfooding

Pass

Audited by Gen Agent Trust Hub on Apr 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for repository maintenance and quality assurance, and its behavior is consistent with its stated purpose.\n- [COMMAND_EXECUTION]: The skill uses shell scripts to automate testing and reporting tasks.\n
  • Evidence: The scripts scripts/generate-dogfooding-report.sh and scripts/validate-dogfooding.sh include logic to prevent directory traversal by checking output file paths for '..' or leading slashes.\n- [EXTERNAL_DOWNLOADS]: The configuration setup for continuous validation references the official @vercel/skills CLI tool.\n
  • Evidence: scripts/setup-continuous-dogfooding.sh generates workflow configurations that install the package from the official registry of a well-known organization (Vercel).\n- [DATA_EXFILTRATION]: The skill provides an optional feature to send status notifications via webhooks using curl.\n
  • Evidence: This functionality is standard for CI/CD status reporting and is designed to use repository secrets for configuration, posing no inherent risk of unauthorized data exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 8, 2026, 02:46 PM