dedup-playbook

Track: Deduplication

Task: Scan the codebase for repeated logic, copy-pasted functions, duplicated condition trees, repeated transformation code, and redundant abstractions.

Goal: Reduce unnecessary duplication only where doing so clearly improves maintainability without obscuring intent.

Rules specific to this track:

• Do not merge code that only looks similar but serves different domain purposes.
• Do not introduce generic helpers that make the code harder to read.
• Consolidate only when the shared behavior is genuinely the same.
• Prefer local duplication over bad abstraction.
• Be especially cautious with UI code, validation logic, business rules, and error flows that may appear similar but diverge subtly.
• If shared logic is extracted, choose names that explain intent, not mechanics.

High-confidence examples:

• truly identical helper logic repeated across files
• repeated constant maps or literal unions that have drift risk
• duplicated utility functions with identical semantics
• repeated small transformation blocks that are clearly the same operation

Avoid:

• merging similar-looking branches with slightly different business meaning
• extracting one-off helpers used in only two places if readability gets worse
• refactors whose main benefit is cleanliness rather than reduced maintenance risk

Output structure:

1. Critical assessment
2. High-confidence / low-risk candidates
3. Medium-confidence items not implemented
4. Low-confidence items untouched
5. Files likely affected
6. Validation plan
7. If editing was requested by the lead, implemented changes and validation results