What problem does it solve? Enabling Native AOT or trimming on a .NET project surfaces dozens of IL analyzer warnings (IL2026, IL2070, IL3050, etc.) caused by reflection-based code that the trimmer cannot statically verify. This Skill provides a systematic, warning-driven workflow to fix every warning without suppressing them incorrectly. ## Core Features & Use Cases - Warning-driven fix loop: Enable IsAotCompatible, build, triage warnings by code, and apply targeted fix recipes iteratively until zero IL warnings remain. - Four fix strategies: Add [DynamicallyAccessedMembers] annotations, refactor broken annotation flow, batch-migrate JsonSerializer calls to source-generated JsonSerializerContext, or mark fundamentally incompatible methods with [RequiresUnreferencedCode]. - Multi-targeting support: Includes polyfills for DynamicallyAccessedMembersAttribute on netstandard2.0 and net472, plus TFM-conditional IsAotCompatible configuration. - Use Case: After upgrading a library to net8.0, you get 40 IL2026 warnings from JsonSerializer calls. The Skill guides you to create one JsonSerializerContext and batch-update all call sites, then cascade annotations outward for the remaining warnings. ## Quick Start Ask the AI to make your .NET project AOT-compatible by fixing all IL trim warnings in the specified .csproj file.