The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides documentation for building API routes that process untrusted external data.
Ingestion points: Data is ingested via request.json() (in the AI proxy example) and url.searchParams (in the weather proxy example).
Boundary markers: The code examples do not include explicit delimiters or instructions to ignore embedded commands within the interpolated variables.
Capability inventory: The skill uses fetch for network operations to external services.
Sanitization: The skill's rules explicitly state to 'ALWAYS validate and sanitize user input,' though the simplified code examples do not demonstrate the specific implementation of these safeguards.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the eas-cli tool and incorporating the @libsql/client library into the project for database connectivity.

api-routes