As AI tools become more common in .NET development, Microsoft emphasizes the growing importance of human oversight. In its latest blog post, the company outlines practical guidance for developers to effectively review AI-generated code—ensuring it remains reliable, maintainable, and aligned with project standards.

Here are the key areas you need to review:

API Design & Interface Architecture
Avoid unnecessary abstraction layers, ensure consistent naming conventions, verify API visibility (public vs internal), and confirm builder extensions follow established conventions.
Testing & Testability
Aim for full coverage of new public methods, prefer snapshot testing over generic assertions, check for overly specific test values, and prevent unwanted changes to existing tests.
File Organization & Architecture
Watch for accidental changes in auto-generated files, confirm code fits within the right architectural context, and keep new classes and interfaces properly organized.
Error Handling & Edge Cases
Apply consistent null-checking, use appropriate exception types and handling strategies, and thoroughly test for edge cases and defensive programming scenarios.
Configuration & Resource Management
Review resource creation, configuration, and cleanup; ensure adherence to established callbacks and configuration patterns; and confirm consistent behavior across contexts.
Code Quality & Standards
Provide full XML documentation for public APIs, maintain clean formatting and style, and assess performance implications of AI-generated designs.

Empower your Dynamics 365 and Power Platform with AI. Learn about Infinity AI Agents and contact us today.