addyosmani/agent-engineer

Agent Engineer - a practical course for software engineers

What it solves

Most software engineers haven't worked with AI agents before, and the topic can feel abstract or tied to fast-changing frameworks. This project solves that by offering a structured, platform-agnostic introduction to agent fundamentals, then shows how to apply those ideas using Google Cloud AI tools. It avoids duplicating API docs and instead links to official resources so readers always see current information.

How it works

The course is split into three parts. Part 1 (101) covers core concepts like how agents think, tool use, design patterns, memory, planning, multi-agent systems, agentic RAG, evaluation, and safety — all in a platform-agnostic way. Part 2 (201) moves into practice with Google Cloud AI: Vertex AI and the Agent Development Kit (ADK), including building a first agent, and protocols like MCP and A2A. Part 3 (301) dives deeper into topics like AGENTS.md, MCP internals, agent skills, and orchestrators. Each lesson is self-contained enough to jump into, but newcomers are encouraged to read in order. The course emphasizes analogies first, fundamentals over frameworks, and honest discussion of trade-offs.

Who it’s for

Software engineers who want to understand what AI agents are and how to build them. No prior AI/ML experience is needed, but basic Python knowledge, a Google Cloud account, and familiarity with REST APIs and JSON are expected.

Highlights

  • Three-part structure: fundamentals, building/shipping, and deep dives.
  • 19 lessons covering everything from agent basics to production deployment.
  • Platform-agnostic core concepts, then hands-on practice with Vertex AI and ADK.
  • Links to official docs and codelabs instead of duplicating tutorials that could go stale.
  • Emphasizes mental models and trade-offs, not just framework features.
  • Community contributions welcome via issues and pull requests.

Related

  • Project
    Sixian-Li/plain-backtest

    PlainBacktest is a framework that enables AI agents to translate natural language trading strategies into reproducible, audit-ready backtests with independent ledger reconciliation.

  • Project
    emperorclaw/emperorclaw

    EmperorClaw is an operational control center for companies running teams of AI agents, providing task tracking, shared knowledge, and auditability to ensure agent work is organized and reliable.

  • Project
    chrisdoc/hevy-mcp

    Hevy MCP is an open‑source Model Context Protocol server that exposes the Hevy fitness‑tracking API as 23 AI‑friendly tools (read and write). It can run as a Cloudflare Worker (hosted endpoint) or locally via npm/Docker, letting assistants like Claude, Codex, or Cursor analyse workouts, summarize training, search routines, and even create or update workouts and body‑measurements—all under user‑controlled authentication (API key or OAuth).

  • Project
    Annyfee/agent-craft

    Agent Craft is a systematic open-source teaching project that guides you through building full-stack AI agents in Python, from basic LLM calls to RAG, LangGraph, MCP, and deployment. It provides 15 modular, runnable examples with blog explanations to help you understand not just how to call frameworks but how agents think, decide, and act.