The Agent Skills Directory

[SAFE]: The skill operates within a restricted logical scope, focusing exclusively on business-centric features and user capabilities while explicitly forbidding technical implementation details such as database design or API architecture.
[SAFE]: Filesystem operations are localized to the tasks/ directory, following a structured naming convention for managing PRDs and Architecture Decision Records (ADRs).
[SAFE]: External research (codebase exploration and web searches) is conducted via distinct tool calls to gather market context and existing project patterns, with findings merged for interactive user clarification.
[SAFE]: The skill implements a comprehensive question protocol and utilizes static templates for document generation, ensuring consistency and preventing arbitrary content execution.
[SAFE]: The workflow incorporates human-in-the-loop checkpoints, requiring user approval for product approaches and scope decisions before proceeding to document generation.

create-prd