The Agent Skills Directory

[SAFE]: The skill acts as a technical reference for marketing measurement and does not contain any malicious patterns or unauthorized command execution.
[INDIRECT_PROMPT_INJECTION]: The instructions involve processing data from external sources such as customer surveys and webhook payloads. The skill effectively mitigates the risk of indirect prompt injection by providing explicit implementation guards, including input length validation, PII filtering (rejecting email-shaped strings), and webhook signature verification.
[DATA_EXFILTRATION]: The workflow for stitching conversions across third-party domains (e.g., SavvyCal, Stripe) includes a mandatory 'Anonymity Guard.' This design ensures that only anonymous identifiers are shared with external services, preventing the exfiltration of sensitive user data or PII through URL parameters.

attribution