dotnet-slopwatch

You can detect LLM reward hacking in .NET code changes by running static analysis to identify common slop patterns like disabled tests, warning suppression, and empty catch blocks, receiving actionable guidance for fixes.

LLM reward hacking in C# source files occurs when AI models take unsafe shortcuts to pass checks, such as disabling tests, suppressing warnings, or adding empty catch blocks, which degrades overall code quality.

To find disabled tests and warning suppression in C# projects, run a static analysis check after code modifications to enforce rules against these slop patterns and prompt for correct implementations.

Yes, you can use static analysis to review project configuration and tests in .NET by validating code modifications across ongoing development to ensure changes maintain quality and reject unsafe shortcuts.

You need to run slopwatch on your .NET codebase after every code modification session to validate changes, detect LLM reward hacking patterns, and ensure ongoing development maintains code quality.

A limitation of detecting empty catch blocks in C# is that the analysis operates strictly during code modification sessions for C# projects and related files, requiring a run after changes to provide remediation guidance.