The Agent Skills Directory

[General] (SAFE): The skill contains standard development documentation and code examples for a frontend application. It does not perform any dangerous operations such as command execution, unauthorized file access, or network requests to untrusted domains. All code snippets follow standard React/Next.js and GraphQL patterns.- [Indirect Prompt Injection] (LOW): The skill describes processing URL parameters (searchParams), which are untrusted data. While this is an attack surface in the resulting application, the skill's instructions focus on standard parsing and filtering logic (e.g., split(','), resolveCategorySlugsToIds) rather than executing instructions from the data. Evidence: Ingestion point at searchParams.categories in page.tsx; Capability inventory: Static filtering and GraphQL variable construction in filter-utils.ts.

filtering-system