story-cover — Novel Cover Generator
Give it a book title and pen name; it infers the genre and uses GPT-Image-2 to produce a professional web-novel cover with the title and author name already rendered.
Image & VideoIntermediate★ 5,602⑂ 846AI score 9/10Last updated: Aug 14, 2026
What it does
- Scans the title (and synopsis if needed) to pick one of ten genres — xianxia, western fantasy, historical romance, modern romance, urban, mystery, sci-fi, historical, horror, light novel — with an explicit priority order for ties.
- Builds the English prompt in three layers: text layer (title typography + a carefully specified author-name style with rules, frames, or glow), platform style layer, and scene layer (character, background, lighting).
- Offers 2–3 composition variants (character close-up / dynamic full body / pure scene) and auto-increments
封面_vN.pngso earlier covers are never overwritten; saves the prompt alongside each image. - Exports a platform-exact upload file (e.g. 600×800) via center-crop + resize using ImageMagick or macOS
sips, so the platform never crops off your title. - Prefers Codex CLI's built-in ImageGen (no key needed) and only falls back to the OpenAI-compatible API with
GPT_IMAGE_API_KEY; supports image-to-image when you supply a reference image.
Who it's for
- Self-publishing novelists and web-fiction authors who need covers fast without hiring a designer.
- Editors or one-person presses producing many covers at consistent quality and exact platform specs.
- Anyone who wants a worked example of layered image prompting with reliable in-image text.
Example uses
- "Make a cover for Sword God Returns by Qingyun for a 3:4 platform" → xianxia detected → golden brush calligraphy title, cloud-scroll author plate → 768×1024 render plus an auto-cropped 600×800 upload file.
- "Use this illustration as a style reference but make it urban fantasy" → set
REF_IMAGEand the skill routes to the/images/editsendpoint. - Unhappy with v1 → switch to composition B (dynamic full body), shift the palette, and regenerate as v2/v3 while keeping the prompt history.
· · · Install guide · · ·
Try it now, no install
Paste this into Claude to use the skill without installing anything.
Read the instructions in this file and follow them to help me: https://raw.githubusercontent.com/worldwonderer/oh-story-claudecode/HEAD/skills/story-cover/SKILL.md What I want: (describe your task here)
If Claude can't open the link, open it yourself and paste the contents instead.
↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.
Install in the Claude app (no terminal)
- Download the ZIP with the button below.
- In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
- Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
Install in Claude Code
Let Claude do it — paste this into Claude Code
Install the skill I found on Claude Skill Mart. Copy the skills/story-cover folder from the GitHub repo worldwonderer/oh-story-claudecode into my ~/.claude/skills/story-cover/. When it's done, tell me in one line what this skill can do.
Install with a command instead
git clone https://github.com/worldwonderer/oh-story-claudecode.git /tmp/oh-story-claudecode && mkdir -p ~/.claude/skills && cp -r /tmp/oh-story-claudecode/skills/story-cover ~/.claude/skills/⚠ This is a third-party skill. Check the source repository before installing.
- Open a terminal and clone the repo:
git clone https://github.com/worldwonderer/oh-story-claudecode.git - Copy the skill into Claude Code:
mkdir -p ~/.claude/skills && cp -r oh-story-claudecode/skills/story-cover ~/.claude/skills/ - Make sure
curl,jq, andbase64are installed (brew install jqon macOS). - For the platform-size export step, install ImageMagick:
brew install imagemagick(macOSsipsworks as a fallback). - Only if you are not using Codex CLI's built-in image generation, set your key:
export GPT_IMAGE_API_KEY=sk-...(addexport GPT_IMAGE_BASE_URL=...for a compatible proxy). - Restart Claude Code, then run
/story-coveror say "make me a novel cover", and supply the title, pen name, target platform, and output directory.
View source on GitHub ↗License: MIT