xhs-imagen

What problem does it solve? Creating a Xiaohongshu carousel normally requires separate research, copywriting, pagination, image prompting, and quality review across multiple tools. This Skill turns a topic, article, or rough idea into a complete publishable package—research notes, post copy, storyboard, a 3:4 cover, 9:16 content pages, and a QA report—in one workflow. ## Core Features & Use Cases - End-to-end content pipeline: Researches and fact-checks claims, selects pages by cognitive anchors instead of mechanical splitting, and writes Xiaohongshu titles, body copy, and hashtags. - Six character-and-style visual profiles: Binds character identity, medium, palette, and anti-drift rules (e.g., alpaca-line-art, toolbox-bot-risograph) so illustrations stay consistent across pages. - Dual rendering paths: Prefers an available image-generation model using one character reference sheet, and falls back to a deterministic local SVG-to-PNG renderer when no image tool exists. - Use Case: Ask for a carousel explaining "why RAG is not installing knowledge into a model" and receive a validated project.json, per-page image prompts, a 3:4 cover, six 9:16 pages, publishable copy, and a QA report checking ratios, text accuracy, and character drift. ## Quick Start Ask your agent to use the xhs-imagen skill to research your topic and create a Xiaohongshu post with a 3:4 cover, several 9:16 content pages, and publishable copy, optionally naming one of the six visual profiles.