# AI use cases for Craft Skills (ZSeven-W/craft-skills) Repo: [ZSeven-W/craft-skills](https://github.com/ZSeven-W/craft-skills) | Use cases: 3 | Page: https://dedupedai.com/use-cases/zseven-w-craft-skills/ Craft Skills is a Python collection of research-backed, eval-driven workflow skills that AI agents can install and run for specific creative and design tasks. This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL. - Draw a 4 to 8 panel diary comic with a fixed character: https://dedupedai.com/use-cases/zseven-w-craft-skills/draw-diary-comic-with-licensed-character/ - Generate transparent PNGs with real alpha, not cut-outs: https://dedupedai.com/use-cases/zseven-w-craft-skills/generate-native-transparent-png/ - Edit and generate Qwen-Image pictures from a reference: https://dedupedai.com/use-cases/zseven-w-craft-skills/edit-and-generate-images-with-qwen/ ## How to draw a 4 to 8 panel diary comic with a fixed character with Craft Skills Repo: [ZSeven-W/craft-skills](https://github.com/ZSeven-W/craft-skills) | Difficulty: Intermediate | Time: 45 minutes | Basis: From the README Use recurring-character-diary-comic to make a full-page diary comic of an authorized recurring character. **Who it is for:** A comic creator with a recurring character they own or are authorized to use, who needs a consistent multi-panel diary page. **The problem:** Keeping one recurring character consistent across panels, with fixed direction and readable dialogue, is hard to do by prompt alone. **The outcome:** A full-page diary comic of 4 to 8 panels, produced with a compared page layout and checked at full size and at 25 percent. ### Steps 1. `git clone https://github.com/ZSeven-W/craft-skills.git` 2. `cp -R craft-skills/skills/recurring-character-diary-comic "${CODEX_HOME:-$HOME/.codex}/skills/"` (inferred, check the README) 3. Install the eval dependencies if you want to run the skill's tests: `python3 -m pip install -r evals/recurring-character-diary-comic/requirements.txt` (inferred, check the README) 4. Restart or reload your agent session, then give the agent your character reference and the diary story. ### Prompt to paste into your agent ```text Use the recurring-character-diary-comic skill from the craft-skills repo to make a 4 to 8 panel diary comic of the authorized character , story: . Return the full-page image and the layout you chose. Check that the dialogue matches the script word for word at both the original size and 25 percent. ``` ### Caveat The README limits this skill to existing characters you have authorization to use, so do not use it with characters you do not have rights to. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/zseven-w-craft-skills/draw-diary-comic-with-licensed-character/ ## How to generate transparent PNGs with real alpha, not cut-outs with Craft Skills Repo: [ZSeven-W/craft-skills](https://github.com/ZSeven-W/craft-skills) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README Use native-transparent-imagegen to get PNG or WebP files that carry the model's own alpha channel. **Who it is for:** A game artist, marketer, or product designer who needs transparent sprites or cutouts and has been getting checkerboard backgrounds baked into RGB files. **The problem:** Many generated images look transparent in a preview but ship with a painted checkerboard or a fake background removal. **The outcome:** Each delivered PNG or WebP file is checked as the model returned it, with alpha, fully transparent pixels, and transparent corners confirmed, or the failure is reported. ### Steps 1. `git clone https://github.com/ZSeven-W/craft-skills.git` 2. `cp -R craft-skills/skills/native-transparent-imagegen "${CODEX_HOME:-$HOME/.codex}/skills/"` (inferred, check the README) 3. Restart or reload your agent session so the skill is discovered. 4. Ask the agent for the transparent image, and review the unmodified output file yourself on light and dark backgrounds. ### Prompt to paste into your agent ```text Use the native-transparent-imagegen skill from the craft-skills repo to generate a transparent PNG of . Keep the original file bytes and verify the alpha channel, fully transparent pixels, and transparent corners. If any check fails, report the failure instead of using a cut-out or color key. ``` ### Caveat The README notes that passing the metadata checks still requires a visual review of hair, glass, smoke, and soft edges on dark and light backgrounds. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/zseven-w-craft-skills/generate-native-transparent-png/ ## How to edit and generate Qwen-Image pictures from a reference with Craft Skills Repo: [ZSeven-W/craft-skills](https://github.com/ZSeven-W/craft-skills) | Difficulty: Advanced | Time: 1 hour | Basis: From the README Use qwen-image-gen to rewrite prompts, edit regions, keep one character across a series, and check the original file. **Who it is for:** A creator who already runs a Qwen-Image setup and needs portraits, edits, or a same-character series from a reference photo. **The problem:** Qwen-Image results often drift off the brief, and sizes and reference relationships are easy to get wrong. **The outcome:** A set of generated or edited images with the prompt rewrite, reference relationship, and size settings recorded, each checked as an original file. ### Steps 1. `git clone https://github.com/ZSeven-W/craft-skills.git` 2. `cp -R craft-skills/skills/qwen-image-gen "${CODEX_HOME:-$HOME/.codex}/skills/"` (inferred, check the README) 3. Restart or reload your agent session, and confirm your existing Qwen connection is available to the agent. 4. Ask the agent for the series, and review each output file at full resolution. ### Prompt to paste into your agent ```text Use the qwen-image-gen skill from the craft-skills repo to create three separate portraits of the person in , one action each, using my existing Qwen connection. Return the three original image files and the rewritten prompt for each. Verify that each file keeps the same identity and the requested frame size. ``` ### Caveat The README says the sample images and offline tests do not guarantee every generation meets the brief, and that a complex chair pose still failed in its comparison. Risk, credentials: The skill runs on your existing Qwen connection, so keep access keys in your own setup and out of prompts and shared files. Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/ URL: https://dedupedai.com/use-cases/zseven-w-craft-skills/edit-and-generate-images-with-qwen/