@slfcys

slfcys-xhs-expert generates xiaohongshu content across three modes (product seeding, lifestyle, rewrite) using conversational tone and scene-specific details. delegates detailed rules to external reference files.

generates video prompts using a five-dimension thinking framework and seven-element video spectrum system. supports single and segmented output modes with consistency mechanisms for multi-segment videos, but lacks concrete procedural detail and failure handling.