Create animated README banners with AI
Explain a repository through a looping SVG hero, keeping one editable source for light and dark versions and reviewing it at README and phone widths.
Create animated README banners with AI around the result a visitor needs to understand. The jano skill reads the project, reduces its purpose to a plain sentence and draws a scene that explains it. Existing identity and supported facts guide the design rather than a fixed promotional template.
An editable source produces separate light and dark SVGs. The headline stays still while the scene loops, and the builder prepares a preview at repository and phone widths. Existing README prose, badges and links remain outside the change.
What it does
- Explain a project through its visible result
- Draw an input-to-output scene for a converter
- Produce light and dark heroes from one source
- Reuse an existing project mark and palette
- Review typography at phone width
- Keep source artwork for later revisions
Install and use
- Supply the complete folder, fonts and builder to a file-capable agent. Use Python 3.9+ with FontTools/Brotli, or the documented uv route, and a browser/image viewing tool for review. Invoke the skill explicitly where the host requires it.
- Try: “Make a README banner for our note-to-report tool. Show notes becoming a report, keep our existing mark and leave the README text alone.” Provide authorised project files.
- Choose one visual explanation and author the SVG. Use only figures backed by the repository; an unmeasured project needs no invented performance claim.
- Build both themes and inspect their lettering, edges and complete loop. Replace the introductory banner only; committing changes is a separate action.
FAQ
Source
Made by jjimenezc (ilien-dev). Original skill: jano, published on GitHub under the MIT licence.
View on GitHub