baoyu-comic

Generates multi-page knowledge comics from source content using prompt-based image generation.

Updated Jun 5, 2026
One-click install
npx skills add https://github.com/xu1713/openhorse --skill baoyu-comic-xu1713
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: baoyu-comic
Source: https://github.com/xu1713/openhorse/tree/main/openhorse/openhorse/skills/creative/baoyu-comic
Command: npx skills add https://github.com/xu1713/openhorse --skill baoyu-comic-xu1713

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Turning educational content, biographies, or tutorials into engaging visual comics requires coordinating storyboarding, character design, style selection, and image generation, which is slow and inconsistent when done manually. ## Core Features & Use Cases - Structured Comic Pipeline: Analyzes source content, builds a storyboard and character definitions, writes per-page image prompts, then generates and downloads each page as a PNG. - Style System: Combines 6 art styles, 7 tones, 7 layouts, and 5 presets (ohmsha, wuxia, shoujo, concept-story, four-panel) with a content-signal auto-selection matrix. - Partial Workflows: Supports storyboard-only, prompts-only, images-only, and per-page regeneration modes for iterative refinement. - Use Case: Paste a technical tutorial and ask for an ohmsha-style educational manga; the skill produces analysis.md, storyboard.md, character definitions, prompt files, and downloaded page PNGs under comic/{topic-slug}/. ## Quick Start Create a knowledge comic from this article about neural networks using the ohmsha preset in portrait format.

Frequently Asked Questions about baoyu-comic

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create an educational comic from text content?

Provide your source content as pasted text, a file path, or a URL and request a knowledge comic. The skill analyzes the content, confirms style and audience with you, builds a storyboard and character definitions, writes per-page prompts, then generates and downloads each page as a PNG.

What art styles and tones are available for comic generation?

Six art styles (ligne-claire, manga, realistic, ink-brush, chalk, minimalist) combine with seven tones (neutral, warm, dramatic, romantic, energetic, vintage, action). Five presets like ohmsha and four-panel add special rules, and a content-signal matrix recommends combinations automatically.

Can I regenerate only specific pages of a comic?

Yes, the regenerate option targets specific pages such as page 3 or pages 2, 5, and 8. It reuses existing prompt files, regenerates only those images, and overwrites the corresponding PNGs after backing up the originals.

Does the image generation accept reference images for character consistency?

No, the image_generate tool is prompt-only and cannot accept reference images. Character consistency is enforced by embedding text descriptions from characters/characters.md inline in every page prompt, while the optional character sheet PNG serves only as a human review artifact.

Why did my generated comic pages not appear in the output directory?

The image tool returns a URL rather than a local file, so each page must be downloaded with curl using an absolute output path. Using relative paths with a drifting shell working directory can silently write PNGs to the wrong location.