How tingaicompass/ai-compass works
Project Type and Purpose
• AI-Compass is an open-source, static documentation/learning platform rather than a dynamic full-stack application.
• It aggregates and organizes AI resources (theory, frameworks, applications, tools, courses, community links) into a navigable ecosystem.
• Users browse a static website (hosted via GitHub Pages/Gitee) built from Markdown content.
File Structure Analysis
• Top-level folders (e.g., “0.AI导航工具集”, “1.1 Prompt工程”, … “10.阿里开源”) each correspond to a major module of content.
• Each module folder contains a single Markdown file and local assets (e.g., images under “picture”).
• Supporting files:
• README.md / README-EN.md – landing page content and navigation.
• weeklyHighlights/ – time-series content for “weekly highlights.”
• No code directories (src, backend, frontend) indicate this is a purely content-centric site using Markdown.
README Insights
• Defines seven core modules:
1. Foundation (基础知识)
2. Tech Frameworks (技术框架)
3. Applied Architectures (应用实践)
4. Products & Tools (产品与工具)
5. Learning Resources (学习资源)
6. Community & Platforms (社区与平台)
7. Enterprise Open Source (企业开源)
• Mentions navigation, search, and external links (GitHub, Gitee, social media).
• Indicates static-site behavior but doesn’t explicitly call out a static site generator (e.g., Jekyll, Hugo).
a. Main Components
• Content Repository (Markdown files + images)
• Static Site Generator (SSG) – transforms Markdown → HTML/CSS/JS
• Hosting (GitHub Pages/Gitee Pages)
• CDN / Static Asset Delivery (for images, CSS, JS)
• Site Navigation & Search (built-in or third-party search widget)
• External Integrations (GitHub, Gitee, social media badges)
b. Relationships and Interactions
• Authors commit Markdown into the Git repo → SSG build pipeline → generated HTML.
• Hosting serves HTML and static assets to end users.
• Navigation component reads site structure (folder tree) to produce menus.
• Search widget indexes generated pages or uses GitHub search.
• External badges/links embed at build time or runtime.
c. Architectural Patterns
• Content-centric, docs-as-code approach.
• Modular directory structure mapping to site navigation.
• Static site generation (no runtime backend).
d. Relevant Technologies
• Git / GitHub / Gitee (version control & hosting)
• Markdown (+ front-matter if any)
• Static Site Generator (e.g., Jekyll, Hugo, MkDocs – choose based on familiarity)
• CDN for picture/asset delivery
• Optional search service (Algolia DocSearch, Lunr.js)