reinforce-skills

SUSPICIOUS. The skill's capabilities mostly match its stated purpose of reinforcing agent memory, and the only executable dependency is the official Vercel/Next.js codemod via npm. However, it persistently rewrites CLAUDE.md with forceful behavioral directives, scans user-level Claude skill/plugin directories, and uses mutable remote npx execution, making it higher-risk than a normal documentation helper even without clear exfiltration or credential theft.