Every MCP server is two things at once: text your model will read, and code that acts with someone's credentials. This guide covers the current MCP specification and its OAuth 2.1 authorization model, five risk classes, a vetting checklist, where Microsoft's MCP hosting and gateway options stand in September 2026, and an allow-list model that holds up in an audit.
Once an AI agent can write to systems, send email or move money, "a human approves it" is not a design. This guide classifies agent actions by consequence and reversibility, maps each class to an approval pattern, and sets out what the audit record must hold to survive a review, grounded in what Microsoft 365 Copilot, Copilot Studio, Microsoft Foundry and Purview actually provide as of September 2026.
Once you have decided to extend Microsoft 365 Copilot, there are three ways to do it, and they are not interchangeable. Copilot connectors change what Copilot knows, declarative agents change how it behaves, and custom engine agents replace its engine. This guide compares the three on licensing, governance and who carries the risk, with every claim checked against Microsoft's documentation in September 2026.