illo skill
An open-source agent skill that turns ideas and articles into original print-style editorial illustrations starring recurring mascot characters.
At a Glance
About illo skill
illo skill is an open-source Agent Skills-format skill made by Trevin Chow. It turns a short idea, a social post, or a full article into original editorial illustrations in which a recurring mascot performs the idea. It runs inside the user's own coding agent rather than as a hosted service.
What It Is
illo skill is an agent skill, not a web app or hosted image API. You invoke it with the /illo command inside an agent such as Claude Code, Codex, Grok CLI, Grok Bot, Hermes, or OpenClaw. It produces illustrations from a bare idea, a URL (for example a hero image plus inline art for a post), or a "surprise me" prompt. The site lists Blot as the built-in character, with community character packs adding more.
Output Formats
Editorial scenes are the default: one caught moment that conveys the idea. Other formats are available on request:
- Mini-comics of 2-4 panels in a single image, for story beats such as fail-to-fix.
- Explainer diagrams with labeled stages, flows, timelines, or stacks.
- Transparent PNG character cutouts of the mascot alone, for slides, docs, or overlays.
- Banner sizes such as an X Article banner, plus custom or brand palettes.
Characters and Looks
Each character pack pairs one mascot with one bundled print look. The site lists 58 packs and 17 looks, including riso, blueprint, bricks, bloom, and snes. Packs install by name, for example /illo install mole character pack. Users can also build their own character, which stays a local folder unless they choose to share it in the community repository.
Setup and Image Backends
The skill needs a coding agent with a capable reasoning model, because it shows each rendered image back to the model to check it and re-roll if needed. Image generation goes through one of several backends: a signed-in Codex CLI or Grok CLI, Grok Bot's built-in image tool, Muse's built-in tool, or an OpenRouter key stored locally. Grok CLI and Grok Bot paths cannot make transparent cutouts, so those route to Codex or OpenRouter. Installation uses each agent's native plugin or skill manager, or npx skills add tmchow/illo-skill --skill illo. The engine is a single stdlib Python script, and the skill and community packs are MIT-licensed.
Community Discussions
Be the first to start a conversation about illo skill
Share your experience with illo skill, ask questions, or help others learn from your insights.
Pricing
Open Source
The illo skill itself is free and open source under the MIT License. Image generation runs through external backends (Codex CLI, Grok CLI, Grok Bot, Muse, or OpenRouter) that carry their own usage costs.
- MIT License
- Works in Claude Code, Codex, Grok CLI, Grok Bot, Hermes, and OpenClaw
- Free with Codex CLI, Grok CLI, or Grok Bot, drawing on that plan's usage with no per-image charge
- Optional OpenRouter backend billed by OpenRouter (BYOK), around a nickel per image (6–25¢ by model)
- Paid OpenRouter fallback only with --allow-paid-fallback
Capabilities
Key Features
- Editorial scenes from a short idea
- Illustrate a whole article from a URL with hero plus inline images
- Mini-comics with 2-4 panels in one image
- Explainer diagrams with labeled stages, flows, and timelines
- Transparent PNG character cutouts
- Surprise me mode that invents a seed and caption-ready line
- 58 installable character packs across 17 print looks
- Build your own private character
- Custom or brand palettes and banner sizes
- Self-check loop that shows each render back to the model and re-rolls
- Multiple image backends including Codex CLI, Grok CLI, Grok Bot, Muse, and OpenRouter
