The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents features that ingest untrusted data from external sources and local files, creating a potential attack surface for indirect prompt injection during the static site generation process.
Ingestion points: references/features-data-loading.md and references/features-dynamic-routes.md describe loading content and metadata from external APIs via fetch and local files (Markdown, CSV) via fs.readFileSync into the site build context.
Boundary markers: The documentation does not mention specific boundary markers or instructions to ignore embedded commands within the processed data.
Capability inventory: The skill utilizes file system reading (fs.readFileSync in references/features-data-loading.md and references/features-dynamic-routes.md), network access (fetch in references/core-config.md), and dynamic template rendering.
Sanitization: No explicit sanitization or validation of the ingested external content is described in the reference materials.

vitepress