dev-ai-slop-cleaner

To remove AI-generated code bloat, scan recent diffs to identify single-use helpers, over-abstractions, generic names, TODO/HACK comments, and unrequested configurability, then propose behavior-preserving simplified replacements. This process strips unnecessary additions to produce concise, maintainable code.

AI code slop consists of single-use helpers, over-abstractions, excessive configurability, and contextless TODO/HACK comments left by LLM patches. This bloat increases technical debt by making code harder to read and maintain, requiring behavior-preserving simplified replacements to resolve.

Refactor multi-file LLM patches by scanning for signature bloat patterns like generic naming and unrequested configurability. The cleanup process proposes simplified, behavior-preserving replacements and prepares actionable patches for executors to apply without regressions.

Yes, you can clean automated executor diffs by proposing behavior-preserving simplified replacements for detected bloat patterns. The process prepares actionable patches and produces verification reports to confirm no behavioral regressions occur during the code cleanup.

The best way to deslop dev-autopilot outputs is to scan the generated bursts for five signature bloat patterns and propose human-reviewable edits. This coordinates handoff to batch cleaners for application and verification, turning noisy AI output into maintainable code.

You should avoid automated code cleanup when recent changes lack distinct bloat signatures like single-use helpers or stray TODO comments. Since the tool proposes behavior-preserving replacements, manually verifying complex logic is necessary if the AI patch introduces intricate, necessary abstractions.