Microsoft Launches .NET MCP Agent Governance Toolkit to Tighten Security Controls

The Model Context Protocol allows AI agents to connect with external data and tools, but it can also expose workflows to tool poisoning, prompt injection and unauthorized actions. Microsoft has therefore brought governance capabilities to the .NET development environment, helping businesses enforce consistent security policies before models invoke tools and reducing the risks of sensitive-data leaks and unauthorized operations.

Microsoft recently released a public preview of its Agent Governance Toolkit MCP Extensions for .NET. Developers can set allow, block and content-filtering rules for AI-agent tool calls and add policy checks between models and external tools. As of July 20, 2026, the product remained in preview, and Microsoft had not announced a general-availability date or pricing.