The training is divided into several technical modules. It begins with an introduction reviewing the fundamentals to prepare the transition to advanced use cases. The program continues with an exploration of interaction modes, to understand the balance between control and autonomy when managing execution through plans and follow-up questions.

Next, a core section is entirely dedicated to AI agents: participants will learn to configure specific behaviors (.agent.md files) to delegate targeted tasks, while grasping the nuances between fast generation models and complex reasoning models. The rest of the program covers extensions and the Model Context Protocol (MCP), enabling integration with external systems and building your own MCP server.

The course also addresses modernization, where the approach taught in this training is systematized through target documentation generation and migration plans.

This training will let you go even further by creating a shared Prompt Registry and structuring documentation specifically designed to guide your agents (Copilot-ready).

You will also master GitHub Copilot Spaces for persistent project context, facilitating onboarding of new employees and management of recurring tasks, as well as MCP architecture to automate complex actions.

Finally, the training concludes with cost and resource management: choosing models to reduce token consumption, governing agent usage, and using Plan Mode and Skills to scope MCPs and formulate precise requests.

Contact us to schedule a GitHub Copilot training tailored to your team’s needs, delivered by a company specializing in DevOps training.