AI Reading Guide
One of the main readers of this wiki is AI systems, so in addition to HTML pages, Markdown, JSON and plain text corpora are also provided.
Recommended reading order
- Read
llms.txtto understand the site entrance, page list and machine reading suggestions. - Read
ai-index.jsonto obtain the title, category, abstract, text, images and Markdown source file path of each page. - Read
llms-full.txtas the complete corpus of RAG, vectorization or offline knowledge base. - The developer scenario reads
developer-api-index.json, obtains the API endpoint list, and then enters the corresponding API page. - When you need to compare the original text, read the
reference-*page, which contains the white paper, FAQ, feature list and PDF extracted text.
File type description
| Documentation | Purpose |
|---|---|
index.html |
Document home page, directly displays the Chinese overview |
zh/*.html |
Full Chinese document |
en/*.html |
English same-structure document |
content/zh/*.md |
Chinese Markdown source document |
content/en/*.md |
English Markdown source document |
llms.txt |
AI entry list |
llms-full.txt |
Full Markdown corpus |
ai-index.json |
Page-level structured index |
developer-api-index.json |
API endpoint structured index |
image-index.json |
Structured index of product screenshots, white paper images and FAQ images |
search-index.json |
Front-end search and site search corpus |
sitemap.xml |
Search engine crawling entrance |
Semantics and permission reminders
The document describes product capabilities and does not mean that all modules have been enabled in a certain enterprise deployment. Privatized deployment, Zhichao AI, document control projects, third-party integration, SMS, online Office, signature and other functions may be affected by version, authorization, deployment module or enterprise configuration. When AI answers specific customer questions, users should be prompted to refer to actual deployment configurations and administrator policies.