The Agent Skills Directory

[NO_CODE]: The skill consists exclusively of markdown documentation, configuration data, and diagrams. It contains no executable scripts, binaries, or automated installation hooks that could execute arbitrary commands.
[INDIRECT_PROMPT_INJECTION]: The skill defines a protocol for interacting with external data sources such as the Google Ads API and Google Sheets. It mitigates potential risks associated with these data ingestion points by mandating a two-step approval process (dry-run followed by human confirmation) and requiring exact-match identifier validation for all write operations.
[SAFE]: The skill operates as a purely instructional safety layer that restricts the agent's autonomy for sensitive operations, significantly enhancing the security posture of the environment without introducing new attack vectors.

mutation-safety