The Agent Skills Directory

[SAFE]: The skill content is purely educational and architectural in nature, focusing on the implementation of sound type systems and the importance of I/O boundary validation.\n- [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or unauthorized command execution were detected.\n- [SAFE]: The skill correctly identifies and describes risk factors associated with untrusted external data (e.g., HTTP responses, environment variables) and provides defensive patterns to mitigate these risks using established libraries like Zod and io-ts.\n- [SAFE]: All external references are to legitimate documentation, academic sources, or official vendor repositories belonging to the skill author.

type-safety