The Agent Skills Directory

[SAFE]: The skill instructions define a robust and secure workflow for interacting with advertising platforms. The logic follows security best practices by enforcing a '--draft' default and requiring explicit user approval for any '--apply' operation. It provides transparency through deterministic mutation plans that detail the blast radius and learning impact before execution. Additionally, the skill mandates policy reviews and preserves immutable audit logs and rollback records, which significantly mitigates risks associated with unintended changes to remote systems. No malicious patterns, obfuscation, or unauthorized data access were identified.

ads-launch