The Agent Skills Directory

[SAFE]: The skill is entirely instructional and provides code templates, design patterns, and validation rules for improving application reliability.
[PROMPT_INJECTION]: The skill uses grounding instructions to ensure responses are based on provided reference files. No attempts to override agent safety protocols or bypass system constraints were identified.
[EXTERNAL_DOWNLOADS]: Code examples reference well-known and trusted libraries such as zod for validation, pino for structured logging, and @sentry/nextjs for error tracking. These are standard professional tools and do not represent a security risk.
[DATA_EXFILTRATION]: There are no commands or patterns that access sensitive local files or attempt to send data to untrusted remote servers.

error-handling