The Agent Skills Directory

[SAFE]: The skill provides robust defensive patterns for credential management, specifically instructing users to store Sanity tokens in private environment variables and Nuxt's private runtimeConfig to prevent them from being leaked to the client bundle (see rules/core-module-setup.md and rules/core-server-routes.md).
[PROMPT_INJECTION]: The skill documents an attack surface for indirect prompt injection through external data ingestion. 1. Ingestion points: Content is fetched from Sanity CMS via useSanityQuery (documented in rules/core-composables.md). 2. Boundary markers: Not explicitly specified for LLM-consumed data. 3. Capability inventory: The skill enables rendering Portable Text, generating dynamic sitemaps, and powering visual editing. 4. Sanitization: The skill actively mitigates risks by recommending validateSanityQuery to prevent query injection in server routes and stegaClean to sanitize slugs in sitemaps (see rules/core-server-routes.md and rules/features-sitemap.md).

nuxt-sanity