The Agent Skills Directory

[PROMPT_INJECTION]: The skill identifies patterns for processing untrusted data from content collections and external APIs, creating a surface for indirect prompt injection.
Ingestion points: Data is ingested via getCollection() and fetch() from remote endpoints as described in references/content-collections.md and references/ssg-patterns.md.
Boundary markers: The provided patterns do not include explicit delimiter-based isolation or instructions to the agent to ignore instructions embedded in the external content.
Capability inventory: The skill demonstrates capabilities for network requests via fetch and potentially unsafe HTML rendering via the set:html directive in SKILL.md and references/ssg-patterns.md.
Sanitization: The examples do not explicitly demonstrate sanitization or validation logic for the content retrieved from external sources before processing or rendering.

astro