A Chinese bootcamp that refuses to stay a bootcamp

What it does

This is a comprehensive Chinese-language learning path for building AI agents, structured like a vocational course but published as a public repo. It covers the full stack from prompt engineering and low-code platforms (Coze, Dify) to code-first frameworks (LangChain, LangGraph, MCP, A2A) and deployment tooling (Docker, Ollama, vLLM). Two complete projects are already live: an e-commerce NL2SQL agent and a multi-agent research system called DeepAgents. A separate interview question bank is organized by job-description skill domains rather than trivia.

The interesting bit

The author explicitly positions this against “fragmented posts and paid bootcamps”—the curriculum is designed to be runnable, not just readable. Each case is supposed to include working code, environment notes, and troubleshooting. The interview bank claims to source from actual big-tech interview experiences, which is unusual for an open tutorial. It also tracks the 2026 job market rather than 2023 tech, updating the stack as frameworks evolve.

Key highlights

• Bilingual in practice: Chinese prose, Python code, targeting Chinese cloud providers (Tencent Cloud, Alibaba Cloud, AutoDL)
• Two shipped projects with separate source repos: e-commerce NL2SQL (LangGraph) and DeepAgents multi-agent research
• Interview bank cross-linked to tutorial sections, organized by competency areas from real JDs
• Covers both low-code (Coze/Dify) and code-first (LangChain/LangGraph) paths, plus fine-tuning (LoRA, QLoRA, Llama-Factory)
• Explicitly excludes Java/Spring AI; Python-only for the core framework track
• Claims all cases are manually run through before publication

Caveats

• The “地表最强” (strongest on earth) branding is enthusiastic; actual depth per topic is hard to verify without working through the full curriculum
• Some sections in the README are truncated; the full outline lives in separate markdown files not shown here
• Two projects are complete, but the broader curriculum is still evolving—check the changelog for what’s actually landed

Verdict

Worth bookmarking if you’re a Mandarin-speaking developer transitioning into AI application engineering and want a structured alternative to paid courses. Less useful if you need deep reference docs on any single framework; this is breadth-first, depth-second by design. English-only readers should look elsewhere.