The Agent Skills Directory

[SAFE]: No malicious patterns, prompt injections, or obfuscated content were detected in the skill files.
[SAFE]: The code examples for custom authentication implementations follow security best practices, using bcrypt with an appropriate cost factor (12) and implementing safeguards against email enumeration.
[SAFE]: Dependencies listed in the documentation (bcrypt, jsonwebtoken, zod) are standard industry libraries for the tasks described.
[SAFE]: The evaluation criteria provided in the evals/ directory correctly test for security-critical features like token expiry, password hashing, and session invalidation.

customer-accounts