The Agent Skills Directory

[SAFE]: The skill is designed for the legitimate purpose of managing wallet security policies and does not exhibit any malicious patterns such as exfiltration, obfuscation, or unauthorized command execution.- [SAFE]: The skill mandates the use of a specialized tool (wallet_propose_policy) for all actions, ensuring that policies are handled through a defined interface rather than direct agent output.- [SAFE]: Security best practices are integrated into the instructions, such as recommending a default DENY for private key exports and implementing a mandatory user approval and signing flow before policies are applied.- [SAFE]: No external dependencies, remote script downloads, or dynamic code execution mechanisms were identified in the provided file.

wallet-policy