Skip to main content
πŸ€– AI entry

LLM / Skill entry

After build, the site emits model-friendly assets: index txt files, full-content txt bundles, and Markdown copies with MDX imports stripped out. This page also includes the weapp-tailwindcss skill install flow so you can plug it into AI workflows directly.

Open llms-quickstart.txtFetch llms-full.txt

Available index files

  • πŸ“„
    llms-index.jsonGEO index with titles, summaries, keywords, and canonicals
  • πŸ“„
    llms.txtSite-wide document link index following llmstxt.org
  • πŸ“„
    llms-full.txtFull content bundle for offline or single-pass loading
  • πŸ“„
    llms-quickstart.txtEnglish introduction and installation guidance

Markdown copies (examples)

Each doc also ships as plain Markdown with the same route path and a .md suffix. For the full list, see llms.txt.

Install and use the skill

If you want AI to generate setup and troubleshooting steps that follow weapp-tailwindcss best practices inside your product project, install the official skill first:

Install command

npx skills add sonofmagic/skills --skill weapp-tailwindcss

List installable skills

npx skills add sonofmagic/skills --list

Install from the local repo for debugging

npx skills add . --skill weapp-tailwindcss

Suggested prompt (new project)

I have a uni-app CLI project with Vue 3 + Vite, targeting WeChat mini apps and H5. Please give me the minimum working setup following the weapp-tailwindcss skill, including install commands, complete config files, and verification steps.

Docs entry points

Recommended loading order

Online mode

  1. Structured retrieval: read https://tw.icebreaker.top/en/llms-index.json first to get titles, summaries, keywords, and canonicals.
  2. Quick start and setup context: load https://tw.icebreaker.top/en/llms-quickstart.txt.
  3. Need navigation context: read https://tw.icebreaker.top/en/llms.txt first to get section names and links.
  4. Full knowledge base: read https://tw.icebreaker.top/en/llms-full.txt.

Offline / batch mode

  • Download llms-full.txt or package the generated Markdown files together.
  • When vectorizing, split by section and use titles from llms.txt as metadata.
  • For metadata retrieval, use llms-index.json as the first-layer index and fetch article bodies by url afterward.
  • When generating answers, cite the exact link or title so the source stays traceable.

Example prompt

You can read the English weapp-tailwindcss introduction and install guide from https://tw.icebreaker.top/en/llms-quickstart.txt. Please cite the corresponding links in your answer.