The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to build web server applications, which inherently possess a surface for indirect prompt injection via the processing of untrusted external data (HTTP requests, WebSocket messages).
Ingestion points: Untrusted data enters the agent-managed context through route handlers in routes/, global and routed middleware/, and defineWebSocketHandler implementations described in references/routing.md and references/tasks-openapi-websocket.md.
Capability inventory: The skill utilizes powerful capabilities including filesystem access via useStorage and assets (references/cache-storage-database.md), network operations via fetch (references/cache-storage-database.md), and experimental task execution via experimental.tasks (references/tasks-openapi-websocket.md).
Boundary markers: The instructions do not explicitly mandate the use of boundary markers or delimiters for untrusted data processed by the server handlers.
Sanitization: No specific sanitization or validation logic is provided in the handler templates; however, this is consistent with a general-purpose framework configuration skill.

nitro