11ai Blog Builder
Use this as the coordinator for the plugin. Read references/scaffold-map.md, then invoke the subsystem skills in order.
Build order
- Use
$11ai-blog-cmsfor content types, routes, registry, validation, library browsing, publication pages, and post readers. - Use
$11ai-blog-authorsafter the CMS shape exists. Add author metadata/tags, postauthorIds, bylines, author routes, and the Authors library mode. - Use
$11ai-blog-markdown-componentsafter post readers exist. Addreact-markdown, GFM, fenced code, prose components, link/image/embed handling, a demo page/post, and shared heading extraction. - Use
$11ai-blog-content-formatwhen migrating, importing, or normalizing post/publication/author content into the expected 11ai metadata and Markdown shape. - Use
$11ai-blog-content-generationwhen the user asks to create new editorial content, seed posts, demo material, publications, or author profiles. - Use
$11ai-blog-tocafter Markdown heading IDs and extraction exist. Add sticky/collapsible TOC UI, active section logic, and tests. - Optional: use
$11ai-blog-uionly when the user asks for 11ai-style visual polish or a reusable editorial UI pass. It must not change the data contracts.
Rules
- Read the target repository's agent instructions and relevant framework docs before writing code.
- Keep the blog file-backed unless the user explicitly asks for a database or hosted CMS.
- Preserve host route prefixes, theme tokens, Tailwind/shadcn conventions, and image pipeline where possible.
- Keep server-rendered pages as server components; isolate client state in browser controls and TOC components.
- Validate after each major layer: typecheck, lint, unit tests, browser tests, and production build when available.
- If the user asks for exact 11ai recreation, use the scaffold map as the default route/component inventory.
Delivery
Report which subsystem skills were used, files changed, route map, content model, UI variants, tests added, checks run, and any intentional deviations from the scaffold.