Complete Agentic AI Engineering Course Repository

Key takeaways#

• Structured six-week path across major agent frameworks.
• Includes setup guides, notebooks, and project code for hands-on work.
• Best for learning and prototyping; production use needs extra safety and cost review.

What this resource is#

The ed-donner/agents repository supports the Complete Agentic AI Engineering Course. The course is organized as a six-week journey for building and deploying autonomous AI agents with OpenAI Agents SDK, CrewAI, LangGraph, AutoGen, and MCP. It includes notebooks, setup guides, project code, and supporting material for learners who want a structured route through modern agent frameworks.

How to use it#

Follow the six-week structure#

The repository is organized into foundations, OpenAI Agents SDK, CrewAI, LangGraph, AutoGen, and MCP. That progression is useful because it starts with agent basics before moving into multi-agent and protocol-based workflows.

Set up the environment carefully#

The README and setup folders include OS-specific setup notes for Windows, macOS, and Linux. The course also calls out dependency details such as CrewAI version pinning and Windows build tools requirements.

Use the repo as a lab notebook#

The repository is not just reading material. It is meant to be opened in an editor, run locally, and adapted while following the course. Treat each week as a lab and commit your own experiments separately.

Production expectations#

Course code is best for learning. Before using any pattern in a real app, review authentication, secrets handling, model costs, evals, retries, logging, and human approval paths.

Verification notes#

This resource was created from the official public repository: https://github.com/ed-donner/agents. Repository metadata can change quickly, so builders should check the README, license, release notes, and setup files before using the material in production or training workflows.