Install
$ agentstack add skill-epicsagas-velith-book-illustrate ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Interior Illustration Pipeline
Orchestrates the illustration generation process for book chapters.
Prerequisites
- Project must have
PRD.mdand at least one chapter draft indrafts/ STYLE.mdshould exist for art direction (falls back to genre defaults if missing)
Pipeline
Step 1: Generate Illustration Plan
Invoke the illustrator agent:
Read the book's PRD.md, STYLE.md, outline.md, and all chapter drafts in drafts/.
For each chapter, identify 2-4 key visual moments suitable for illustration.
Generate style-consistent illustration prompts with a reusable style seed.
Output the complete plan to publish/illustrations/plan.md and manifest.json.
The agent will:
- Read STYLE.md for art direction (or infer from genre)
- Scan each chapter for illustration-worthy scenes
- Create a style seed prompt for consistency
- Generate prompts for each scene in multiple tiers
- Write the plan and manifest
Step 2: Review Plan
Present the illustration plan to the user for review:
- Total illustration count
- Per-chapter distribution
- Style seed summary
- Key scenes selected
Ask the user to approve, modify scene selections, or adjust the style direction.
Step 3: Execute (Optional — requires image generation tools)
If image generation MCP tools are available (claude-image-gen, Replicate, etc.):
- Generate images using the prompts from the plan
- Save to
publish/illustrations/ch{NN}-{scene-slug}.jpg - Optimize file sizes (compress to <500KB for EPUB compatibility)
If no image generation tools are available:
- The plan.md serves as the executable spec
- User can use the prompts in Midjourney, DALL-E, Stable Diffusion, or any external tool
- Place generated images in
publish/illustrations/following the naming convention
Step 4: Integration
Insert illustration references into chapter drafts:
Placement rules:
chapter-header: After the chapter title headingfull-page: On its own line, between paragraphs at the scene locationinline: Within the paragraph flowspot: As a decorative element at section breaks
Output Structure
publish/illustrations/
├── plan.md # Illustration plan with prompts
├── manifest.json # Machine-readable summary
├── ch01-opening-scene.jpg # Generated images (if executed)
├── ch01-climax.jpg
├── ch02-meeting.jpg
└── ...
EPUB Compatibility
- All images: JPEG format, RGB, sRGB color space
- Target: <500KB per image for reasonable EPUB file size
- Recommended dimensions: 800x1200px (full-page), 1600x533px (header), 600x600px (inline/spot)
- pandoc automatically embeds images referenced in markdown
Post-Completion
Update the project dashboard status:
node {PLUGIN_ROOT}/skills/book-status/scripts/scan-project.js [project-dir] --plugin-root={PLUGIN_ROOT}
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: epicsagas
- Source: epicsagas/Velith
- License: Apache-2.0
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.