deslop

What This Skill Does

The deslop skill is a specialized code-cleanup utility designed to strip away the repetitive, low-value, or overly verbose code patterns often introduced by automated AI code generation. Often referred to as 'AI slop,' this includes excessive defensive comments, redundant type casting, unnecessary catch-all error handling that ignores context, and placeholders that clutter the codebase. The deslop agent acts as a surgical tool, analyzing your git diff against your project's main branch to identify code that differs from your existing human-written stylistic and structural patterns. It meticulously reviews candidate changes, removes noise that doesn't contribute to functional requirements, and ensures your codebase remains clean, maintainable, and idiomatic to your project standards.

Installation

To integrate this skill into your environment, use the OpenClaw CLI: clawhub install openclaw/skills/skills/brennerspear/deslop

Use Cases

Use deslop whenever your PRs start feeling bloated with AI-generated boilerplate. It is particularly effective after a large-scale refactor or feature addition where an AI agent assisted in writing the heavy lifting but left behind excess noise. It is ideal for teams enforcing strict code quality guidelines where readability and conciseness are prioritized over the 'safe-but-ugly' patterns frequently emitted by LLMs. It works best when you want to automate the mundane task of manual code review and cleanup.

Example Prompts

1. "Deslop my latest changes in the authentication module; it's littered with redundant try-catch blocks and excessive logging."
2. "Review the current git diff against main and remove any obvious AI-generated filler comments or weird type casting that violates our local style."
3. "Run a cleanup pass on these files. Remove the defensive programming overhead the AI added, but make sure you keep the critical input validation logic."

Tips & Limitations

To get the most out of deslop, ensure your references/slop-heuristics.md is updated with your project's specific coding standards. The tool is designed to prioritize safety; it will never remove security-critical logic, such as input sanitization, authentication checks, or file I/O protections. Always run your suite of tests—such as bun check or bun typecheck—after a deslop pass to ensure that no functional regressions were introduced. Note that while deslop is highly effective at catching stylistic inconsistencies, it relies on context. The more consistent your existing codebase is, the more accurate the tool will be at identifying 'slop' versus 'essential logic'. If the skill flagged something important, verify the diff and adjust the heuristics to ignore those patterns in the future.