The Agent Skills Directory

[SAFE]: The skill incorporates security guidelines that instruct the agent to use --readonly flags and --dry-run modes where applicable.
[SAFE]: Instructions explicitly require the agent to confirm the account, target object, and specific mutation before executing any write or delete operations.
[SAFE]: No evidence of prompt injection, obfuscation, or unauthorized data exfiltration was detected.
[SAFE]: The skill provides specific commands for Google Meet operations through an established CLI tool ('gog') without attempting to download untrusted external code or elevate privileges.

gog-meet