The Agent Skills Directory

[SAFE]: The skill serves as a reference guide for Sanity development. It contains no executable scripts or dynamic instructions that could compromise the environment.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill consistently recommends using environment variables (e.g., NEXT_PUBLIC_SANITY_PROJECT_ID, SANITY_API_READ_TOKEN) and .env files for managing project credentials and tokens, adhering to secure secret management practices. No hardcoded credentials were found.
[COMMAND_EXECUTION]: Shell commands described in the guides (e.g., npx sanity init, npm install, sanity blueprints deploy) are standard development workflows for initializing projects, managing dependencies, and deploying infrastructure through official Sanity CLI tools.
[INDIRECT_PROMPT_INJECTION]: The skill provides patterns for processing external data from the Sanity Content Lake. It recommends using stegaClean to remove invisible metadata characters from strings used in application logic, which serves as a functional safeguard for data integrity. The attack surface is limited to standard data-fetching patterns in common web frameworks.
[EXTERNAL_DOWNLOADS]: All external references, packages, and code templates target official Sanity repositories, well-known frontend frameworks (Next.js, Nuxt, Astro), or reputable community plugins. These resources originate from trusted sources and are integral to the skill's purpose.

sanity-best-practices