skills/helderberto/skills/perf-audit/Gen Agent Trust Hub

perf-audit

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No evidence of override markers, safety bypass attempts, or malicious role-play instructions was found in the skill or its references.
  • [DATA_EXFILTRATION]: No hardcoded credentials or access to sensitive file paths (such as .env or SSH keys) were detected. The skill reads package.json and build metadata, which is consistent with its stated purpose.
  • [COMMAND_EXECUTION]: The skill executes build and analysis commands (npm run build, npx). These are restricted to local project scripts and standard analysis tools.
  • [EXTERNAL_DOWNLOADS]: The skill uses npx to execute vite-bundle-visualizer and webpack-bundle-analyzer. These are well-known, industry-standard packages from the NPM registry.
  • [REMOTE_CODE_EXECUTION]: No patterns involving piping remote scripts to a shell or executing untrusted code from external URLs were found.
  • [SAFE]: The skill follows its stated purpose without hidden behaviors or security risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 10:23 AM