The Agent Skills Directory

[PROMPT_INJECTION]: Indirect prompt injection surface detected. The skill processes untrusted user descriptions to generate PRD files stored in the tasks/ directory.\n- [PROMPT_INJECTION]: Ingestion point: The skill receives feature descriptions directly from the user which are then interpolated into markdown templates.\n- [PROMPT_INJECTION]: Boundary markers: The skill lacks explicit delimiters or instructions to prevent downstream agents from obeying commands embedded within the user-provided descriptions.\n- [PROMPT_INJECTION]: Capability inventory: The skill performs file-write operations to the tasks/ directory, creating a persistence layer for untrusted input.\n- [PROMPT_INJECTION]: Sanitization: No sanitization or validation of the user's input is performed before it is written to the filesystem.

prd