/oma-deslop

/oma:deslop

Purpose: Clean up low-quality AI artifacts, verbose patterns, over-engineered solutions, and anti-patterns while preserving functionality.

Usage: /oma:deslop <target>

Examples:

• /oma:deslop fix over-engineered code in src/utils/
• /oma:deslop verbose logging patterns across the codebase

---

How It Works

Anti-Slop Detection Patterns

Verbose/Redundant Patterns:

• Commented-out code blocks
• Overly verbose variable names (e.g., the_final_result_string)
• Excessive abstraction layers for simple operations
• Unnecessary type annotations where inference suffices

AI Artifacts:

• Try-this patterns without context
• Uncalled helper functions
• TODO comments left by previous AI turns
• Stale skeleton code

Over-Engineering:

• Factory factories
• Singleton patterns where plain modules suffice
• Overly generic types (e.g., Map<String, Object>)
• Premature abstractions

Process

1. Protect current behavior — run regression tests or establish baseline
2. Scan target for known anti-patterns
3. Flag each issue with severity (HIGH/MEDIUM/LOW)
4. Write cleanup plan — list concrete smells, order from safest deletion to riskier consolidation
5. Fix HIGH and MEDIUM issues automatically
6. Run quality gates — regression tests must stay green
7. Report findings with before/after diffs
8. Preserve all functional behavior — regression-safe

Constraints

• Never break tests — run test suite after changes
• Never remove business logic — only cleanup artifacts
• Ask before touching complex refactors (MEDIUM → ask user)
• Document all changes in the output summary