The Agent Skills Directory

[SAFE]: The skill's operations are confined to the local filesystem. It reads existing diagnostic reports (typescript.json, lint.json) and samples component source code to generate a performance and hygiene report.
[SAFE]: No network activity, data exfiltration, or sensitive credential access was identified in the instructions or metadata.
[SAFE]: The installation methods and external references point to the author's own repositories and official community tools (such as the skills CLI from Vercel Labs), which is consistent with the skill's intended use case.

react-expert