Hello-Agents: Datawhale's Open-Source Systematic Tutorial for Building AI Agents from Scratch
datawhalechina/hello-agents is an open-source AI agent learning tutorial by the Datawhale community, focusing on truly AI-native agent design rather than low-code, flow-driven platforms like Dify or n8n. Its mission is to guide developers from mere LLM users to genuine agent system builders, filling the critical gap in systematic, practice-oriented Agent tutorials. The project has amassed 24,074 stars with 147 new stars per day.
The 16-chapter curriculum spans agent theory (definitions, history, ReAct/Plan-and-Solve/Reflection paradigms), framework practice (AutoGen, AgentScope, LangGraph), context engineering, memory systems, communication protocols (MCP/A2A/ANP), Agentic RL (SFT to GRPO full pipeline), performance evaluation, and end-to-end projects like an intelligent travel assistant and a Cybercity multi-agent simulation. The bundled HelloAgents framework lets readers build their own agent framework from zero using OpenAI's native API.
Ideal for Python developers and AI practitioners who want to understand agent internals deeply. Completely free and open-source under Datawhale — the go-to starting point for the Agent era of 2025.
Hello-Agents: Datawhale's Definitive Open-Source Agent Tutorial
Background
If 2024 was the year of the model wars, 2025 is unquestionably the Year of Agents. The industry's focus has shifted from training larger foundation models to building smarter agentic applications. Datawhale's hello-agents fills the critical gap in systematic, practice-oriented Agent tutorials. The project has reached 24,074 GitHub stars (+147/day).
Curriculum Structure
16 chapters across five parts: (1) Theory — agent definitions, history (symbolism to LLM-driven), and LLM fundamentals; (2) Building Agents — ReAct/Plan-and-Solve/Reflection hands-on, framework practice (AutoGen, AgentScope, LangGraph), and HelloAgents from scratch with OpenAI native API; (3) Advanced — memory/RAG, context engineering, MCP/A2A/ANP protocols, Agentic RL (SFT to GRPO); (4) Capstone projects — travel assistant, deep-research agent, Cybercity simulation; (5) Graduation project.
Why It Stands Out
The tutorial distinguishes flow-driven software engineering agents (Dify/n8n) from true AI-native agents with autonomous planning, memory, and tool use — focusing exclusively on the latter.
Industry Trend
hello-agents reflects the pivot from model competition to application ecosystem competition. MCP adoption, A2A communication standards, and Agentic RL industrialization are all driving agents into production. Mastering agent construction is becoming a core competency for AI engineers worldwide.
In-Depth Analysis and Industry Outlook
From a broader perspective, this development reflects the accelerating trend of AI technology transitioning from laboratories to industrial applications. Industry analysts widely agree that 2026 will be a pivotal year for AI commercialization. On the technical front, large model inference efficiency continues to improve while deployment costs decline, enabling more SMEs to access advanced AI capabilities. On the market front, enterprise expectations for AI investment returns are shifting from long-term strategic value to short-term quantifiable gains.
However, the rapid proliferation of AI also brings new challenges: increasing complexity of data privacy protection, growing demands for AI decision transparency, and difficulties in cross-border AI governance coordination. Regulatory authorities across multiple countries are closely monitoring these developments, attempting to balance innovation promotion with risk prevention. For investors, identifying AI companies with truly sustainable competitive advantages has become increasingly critical as the market transitions from hype to value validation.