The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents how to process untrusted data from web requests (Request objects), path parameters (context.params), and form submissions (formSubmitted event).
Ingestion points: SKILL.md describes handlers that take Request and Context objects, and event handlers for formSubmitted.
Boundary markers: The provided code examples do not include explicit instructions or markers to delimit untrusted data from the agent's instructions.
Capability inventory: The skill demonstrates capabilities such as making network requests (fetch in SKILL.md), executing database queries (db.sql in SKILL.md), writing to storage (getStore().set in SKILL.md), and purging caches (purgeCache in SKILL.md).
Sanitization: The examples do not explicitly demonstrate sanitization or validation of input data before processing, which could allow malicious instructions embedded in web requests to influence downstream operations.

netlify-functions