Agent Skills
› zereight/gitlab-mcp
› ai-slop-cleaner
ai-slop-cleaner
GitHub用于清理AI生成的冗余、重复或过度抽象的代码，采用先删除后重构的工作流，确保行为不变且回归安全。适用于代码膨胀、死码及边界泄漏等场景，通过运行测试验证变更安全性。
Trigger Scenarios
deslop
anti-slop
AI slop
cleanup slop
clean up this code
remove bloat
Install
npx skills add zereight/gitlab-mcp --skill ai-slop-cleaner -g -y
SKILL.md
Frontmatter
{
"name": "ai-slop-cleaner",
"description": "Clean AI-generated code slop with regression-safe workflow. Activate when user says: deslop, anti-slop, AI slop, cleanup slop, clean up this code, remove bloat.\n",
"argument-hint": "[--review] [files or scope]"
}
AI Slop Cleaner
Clean AI-generated code slop without changing behavior. Deletion-first workflow with regression safety.
When to Use
- Code that works but feels bloated, repetitive, or over-abstracted
- Duplicate logic, dead code, wrapper layers, boundary leaks
--reviewfor reviewer-only mode (no changes, just findings)
When NOT to Use
- New feature build or product change
- Broad redesign
- Behavior is unclear and untestable
Principles
- Preserve behavior unless explicitly asked for changes
- Delete before refactoring — removal is safest
- Scope-bounded: only clean what was specified
- Regression-safe: verify tests pass after every change
Slop Signals
- Unnecessary wrapper functions
- Duplicated logic across files
- Over-abstracted single-use helpers
- Dead code / unreachable branches
- Excessive comments restating obvious code
- Premature abstractions (used only once)
Workflow
- Identify target files/scope
- Run existing tests (baseline)
- Apply deletions first (dead code, unused imports)
- Simplify remaining code (reduce nesting, merge duplicates)
- Re-run tests after each change
- Report what was cleaned and verification results
Version History
- a8f5d4f Current 2026-07-25 10:03
