The Agent Skills Directory

[SAFE]: The skill is currently a specification and does not contain executable code or scripts.
[SAFE]: The specification correctly defines security requirements for authentication, explicitly instructing that credentials must be handled via environment variables rather than being hardcoded.
[SAFE]: The skill utilizes official and trusted libraries (google-ads for Python and google-ads-node for Node.js) from a well-known service provider (Google).
[SAFE]: The specification includes robust pre-flight validation steps, such as credential checks and status verification, to ensure operational safety.
[PROMPT_INJECTION]: The skill defines an ingestion point for external data via a configuration file (campaign-spec.yaml), which creates a surface for indirect prompt injection. However, the specification includes validation logic and mandates a security review by a specialized squad before any implementation.
Ingestion points: The campaign-spec.yaml file referenced in the SKILL.md inputs section.
Boundary markers: None are explicitly defined in this specification to isolate the YAML content from the agent's instructions.
Capability inventory: The skill is designed to create and modify Google Ads campaigns, budgets, and assets via API calls.
Sanitization: The specification requires pre-flight validation and verification of conversion actions, though it does not explicitly detail the sanitization of instructional content within the YAML values.

google-ads-api