The Agent Skills Directory

[SAFE]: The skill provides a bundled Python script for interacting with the Cloud Mail API. It handles authentication securely using environment variables rather than hardcoding credentials or prompting the user to share them in chat.
[SAFE]: Token management implements security best practices. Caching is performed in the user's home directory with restricted file permissions (0600), ensuring that only the current user can read the cached authentication token.
[SAFE]: The skill documentation includes clear safety rules for the AI agent, specifically requiring human confirmation for any deletion operation and advising on how to handle user passwords securely.
[SAFE]: Network operations are restricted to the domain provided in the environment configuration. The CLI prohibits insecure HTTP connections to remote hosts by default, protecting credentials from interception over the network.

cmail