GitHub · GH-700 · BETA
Models, skills, prompts, and MCP servers
GH-700 · Lesson 3 of 15 · 17:34
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Lesson 3 of the free GH-700 series (domain 1): how models shape output and explainability, choosing a model (lighter, general-purpose, reasoning, reasoning effort, Auto model selection, the cache), the characteristics of custom instructions, prompt files, and agent skills, and when to use Model Context Protocol (MCP) servers. With three quick checks and a recap.
Preview the topics
- An analyst asks Copilot to reformat a small table, using the most capable reasoning model with a high reasoning effort. What should the analyst change?
- Contoso wants Copilot to follow its brand writing rules in every answer, without anyone asking. Which customization fits?
- Contoso wants Copilot to read customer tickets from its support system when it drafts the weekly report. What should the team add?