awesome-agentic-ai-zh
A trilingual (Traditional Chinese, Simplified Chinese, English) learning roadmap for agentic AI, from LLM basics to multi-agent systems, with curated resources and hands-on exercises.
At a Glance
Learning roadmap and curated resources repository released under the MIT License, free of charge.
Engagement
Available On
Alternatives
Listed Nov 2026
About awesome-agentic-ai-zh
awesome-agentic-ai-zh is an open-source learning map for AI agents, maintained by Wenyu Chiou and published as a MkDocs documentation site in Traditional Chinese, Simplified Chinese and English. It leads readers from "what is an AI agent" through to building reliable agent systems. The latest listed release is v2026.10.29-3, and the project is MIT licensed.
What It Is
The project is a learning roadmap plus curated resources plus small runnable exercises, rather than a software product. The README says it does not rewrite a full encyclopedia; where a full chapter is needed it points to official documentation, Datawhale Hello-Agents or a matching cookbook. Key terms are explained in plain language on first use, and a glossary is provided.
How the Learning Path Is Organized
The map has 10 learning stations: Stage 0 (foundations: Python, Git, APIs), Stages 1-8 and an advanced reading station, Stage 7.5. Stages 0-2 are shared foundations covering LLM basics and prompt design. Readers then choose a track:
- Track A (CLI power user): choosing a CLI agent, building repeatable workflows, and connecting to real work, using tools such as Claude Code, Codex and OpenCode.
- Track B (agent builder): tool use and the first agent loop, workflow graphs and frameworks, the Claude Code ecosystem (MCP, Skills, Plugins, Hooks, Subagents), memory and RAG, production engineering (evals, observability, budgets, human-in-the-loop, recovery) and agent interfaces such as Computer Use and Browser Use.
A separate everyday-user path suits people who want to use AI safely without writing code.
Role-Based Branches and Practice
Beyond the tracks, role guides cover researchers, developers, teachers, knowledge workers and everyday users. The site also offers a progress tracker, a capstone project, a seven-step walkthrough for building a first agent, a cookbook, and runnable examples that can be executed offline.
Setup Path
Readers can read the documentation site online or clone the repository and open the foundations stage. A beginner setup guide and a how-to-use page are included.
Community Discussions
Be the first to start a conversation about awesome-agentic-ai-zh
Share your experience with awesome-agentic-ai-zh, ask questions, or help others learn from your insights.
Pricing
Open Source (MIT)
Learning roadmap and curated resources repository released under the MIT License, free of charge.
- MIT License covers the repository content
- Trilingual (繁中 / English / 简中) learning roadmap
- 240+ curated resources and hands-on examples
- Online documentation site
Capabilities
Key Features
- Trilingual content (Traditional Chinese, Simplified Chinese, English)
- 10 learning stations from foundations to agent interfaces
- Track A (CLI agents) and Track B (agent builder) paths
- Role-based guides for researchers, developers, teachers, knowledge workers and everyday users
- Hands-on exercises and offline runnable examples
- Glossary, cookbook and curated resource list
- Progress tracker and capstone project
- Seven-step walkthrough for building a first agent
