Baoyu Claude Code Toolkit

What Problem It Addresses

Baoyu's Claude Code skill suite is wrapped as an OpenClaw helper for repeatable content workflows. It turns content production into callable tasks: generating Xiaohongshu image posts, cover images, infographics, slide decks, comics, and article illustrations, plus utilities for translation, URL-to-Markdown, X-to-Markdown, image compression, Markdown formatting, and Markdown-to-HTML. Rather than being a single generator, it groups prompt selection, generation backends, document conversion, and publishing actions in one skill directory. It fits engineering workflows that repeatedly produce visual assets, clean web content, or process Markdown documents.

How It Works and Where to Watch

For six content-generation skills — baoyu-xhs-images, baoyu-infographic, baoyu-cover-image, baoyu-slide-deck, baoyu-comic, and baoyu-article-illustrator — the wrapper first opens baoyu-prompt-tool.html and asks the user to select style, layout, color, type, or tone before execution: baoyu-xhs-images uses style and layout, baoyu-cover-image uses type, color, render, text, and tone, baoyu-slide-deck uses a preset, and baoyu-article-illustrator uses type and style. Image generation is delegated to baoyu-imagine, which can call OpenAI, Azure OpenAI, Google, OpenRouter, DashScope, MiniMax, Replicate, and other providers. Publishing skills rely on real Chrome or a browser login for X, WeChat Official Accounts, and Weibo; web capture uses Chrome CDP. Watch that baoyu-danger-x-to-markdown and baoyu-danger-gemini-web require authentication, and the selected parameters directly shape the output. In practice, confirm the prompt page has fully loaded and check that the provider keys are available before running a full batch.