Building AI Agents

Fifteen modules covering AI agent fundamentals — from concepts, prompt engineering, and API setup through RAG, MCP, multi-agent orchestration, safety, and deployment.

AI AgentsLLM · RAG · MCP · A2A15 modules · EN / 中文
Module 01
What is an AI Agent?
Definition, agent vs chatbot, core components, and the 2025-2026 agent landscape.
Module 02
Environment Setup & API Configuration
Python environment, API keys, SDK installation, and first connection.
Module 03
Prompt Engineering
Effective prompts, system prompts, few-shot and chain-of-thought, templates, debugging.
Module 04
Your First API Call
Chat completions, system/user/assistant messages, temperature, tokens.
Module 05
Building the Agent Loop
The observe-think-act cycle, core agent loop, multi-step reasoning.
Module 06
Tool Use & Function Calling
Defining tools, function calling protocols, tool execution.
Module 07
Memory & Context Management
Short/long-term memory, context window strategies, persistent storage.
Module 08
RAG — Retrieval-Augmented Generation
Embeddings, vector databases, document chunking, RAG pipelines.
Module 09
MCP — Building Your Own Servers
MCP servers exposing tools, resources, and prompts to agents.
Module 10
LangChain & CrewAI
Agent frameworks — LangChain for general-purpose, CrewAI for role-based teams.
Module 11
Multi-Agent Orchestration
Agent communication, delegation, supervisor architectures, collaborative workflows.
Module 12
A2A Protocol
Google's Agent-to-Agent protocol for standardised inter-agent communication.
Module 13
Safety & Guardrails
Prompt injection defence, output validation, guardrail frameworks, human-in-the-loop.
Module 14
Testing & Evaluation
Unit testing, evaluation frameworks, benchmarks, measuring performance.
Module 15
Deployment & Best Practices
Production deployment, monitoring, security, cost management.