The Agent Skills Directory

[SAFE]: The skill implements robust security controls for handling sensitive cryptographic keys and financial transactions. Security rules explicitly forbid hardcoding, logging, or displaying private keys and API keys. The code generation patterns consistently use environment variables for secret management.
[SAFE]: A mandatory Decision Guide and explicit confirmation rules ensure no fund movements occur without direct user approval of all parameters. The instruction hierarchy correctly prioritizes safety rules over repository context, mitigating potential indirect prompt injection from codebase files.
[SAFE]: All dependencies and documentation links point to official vendor resources or reputable ecosystem libraries. The skill demonstrates clear alignment between its stated purpose and implemented behavior.

unify-balance