Menu

Individual skill

book-visuals

Book visual system: art bible and look lock, figure plan per chapter, code-rendered diagrams/charts/technical drawings, diffusion-generated illustrations and spot art, photo sourcing policy, cover, model-agnostic prompt compilation, vision QA, and asset validation for print and EPUB. Every image matches the book, the chapter, and the author's brief regardless of which image model renders it.

plugins/epicsagas/Velith/skills/book-visuals/SKILL.md

Read pinned source

Source commit d780783eba4ecc7c8b9429f95c9362df95f07c61

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
Yes
Repository scan complete
No
Skill scan complete
No
Observed behavior
Network or broad shell
Last checked
Nov 4, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
1,532
Repository last updated
Nov 6, 2026
License
Apache-2.0

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

# Book Visuals

Goal: every image in the book looks like it belongs to this book, sits where the text needs it, and is release-ready on the first export. Images that vary with the model, drift between chapters, or decorate instead of doing work are the visual equivalent of AI prose tells.

Read `${CLAUDE_PLUGIN_ROOT}/skills/loom/quality-bar.md` (the visual section) and the genre skill's **Visuals** section.

## Why images go wrong, and the fix

| Cause | Fix in this pipeline |
|-------|----------------------|
| No shared visual spec; each prompt invents a style | `art-bible.md`: one canonical look, palette, medium, composition rules, character and setting constants, negative rules. Every prompt is derived from it |
| Free-form prompts differ per backend | `velith.mjs images compile`: one structured image spec → per-backend prompt syntax (Midjourney, gpt-image, Stable Diffusion/FLUX, Imagen, Ideogram) with the art bible baked in |
| Diagrams and drawings rendered by diffusion models | Never. Diagrams, charts, technical drawings, maps' base layers are **code-rendered** (Mermaid, D2, Graphviz, SVG, matplotlib/Vega) by `figure-engineer`. Text-accurate, editable, print-safe |
| No rejection loop | `art-director` reads every generated image (vision), scores it against the art bible, rejects and regenerates with corrected prompts; reviews the full contact sheet for consistency |