The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from external API services through TanStack Query hooks, creating a surface for indirect prompt injection. * Ingestion points: External data is ingested via queryFn implementations across all hooks, such as useSuspenseQuery and useInfiniteQuery, particularly through the postsApi service pattern. * Boundary markers: Documentation does not demonstrate the use of delimiters or specific instructions for the agent to ignore potential prompts embedded within the fetched API data. * Capability inventory: The skill enables the agent to perform network read and write operations (mutations) and process the results within the application lifecycle. * Sanitization: The patterns shown focus on functional implementation and do not include content sanitization or validation steps to neutralize potential prompt injection attacks present in API responses.

tanstack-query