The Agent Skills Directory

[Official Resource Usage]: The skill exclusively references official Google documentation, GitHub repositories (github.com/googleapis), and package registries for its libraries (e.g., google-ads on PyPI, googleads/google-ads-php on Composer). All URLs and dependencies are from well-known, trusted domains belonging to the vendor.
[Credential Management Best Practices]: The skill guides users to manage sensitive credentials (Client ID, Client Secret, Refresh Token) using local configuration files like google-ads.yaml or ads.properties and environment variables. This aligns with industry-standard security practices for local development and avoids hardcoding secrets.
[Authentication Security]: The skill correctly uses the official Google Cloud CLI (gcloud) and the OAuth 2.0 authorization code flow for generating refresh tokens. It emphasizes using a 'Test User' and 'Test Accounts' when starting, which reduces the risk of accidental production data modification.
[Input Sanitization]: The instructions include mandatory steps for the agent to 'normalize' user-provided data, such as customer IDs, by stripping non-numeric characters before generating code, which acts as a safeguard against malformed input errors.

google-ads-api-quickstart