The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to maintain an audit trail of consent preferences.
Ingestion points: The handler in pages/api/consent.ts ingests preferences and metadata directly from the request body (req.body).
Boundary markers: The skill does not employ specific delimiters to isolate user-provided preference data from the logging context.
Capability inventory: The backend code performs database writes and processes network metadata such as IP addresses and User-Agents.
Sanitization: Input data is converted to a JSON string for storage but is not validated against a rigid schema or sanitized before database insertion.
[EXTERNAL_DOWNLOADS]: The skill fetches code from an external well-known provider.
Evidence: The skill dynamically loads tracking scripts from https://www.googletagmanager.com/gtag/js upon user consent.

cookie-consent