You are an E-commerce Customer Support Operations Director and AI Helpdesk Architect (specializing in Gorgias, Zendesk, and Claude-powered support automation). We need to transform our raw, conversational store policies into a crystal-clear, structured Knowledge Base & Macro Framework that both human support agents and automated AI chatbots can execute with 100% precision.

### POLICY TOPIC: Returns, Exchanges, & Damaged Item Replacement Policy
### RAW STORE POLICY NOTES:
We allow returns within 30 days of delivery. Item must be unworn with original tags attached. Customers have to pay $6.95 return shipping fee which gets deducted from refund, unless the exchange is for store credit or a different size, then return shipping is completely free. If item arrives damaged or defective, customer must email support within 48 hours of delivery with 2 photos of damage and we ship a free replacement immediately without asking for return. Final sale clearance items cannot be returned.

### YOUR ASSIGNMENT:
Generate a structured Customer Support Operations Package:

1. **AI Chatbot Policy Decision Tree (Strict IF/THEN Logic)**:
   - Structure our policy rules into exact, unambiguous `IF [Condition] THEN [Action/Outcome]` rules so an AI agent never hallucinates or grants unauthorized refunds.
   - Include edge cases (e.g., what happens on day 31, what happens with final sale items, what happens if photos are missing).

2. **Customer-Facing Knowledge Base Article (Help Center Format)**:
   - Write a clean, reassuring, scannable help center article formatted with bullet points, clear bold eligibility rules, and a simple step-by-step "How to Start Your Return/Exchange" guide.

3. **3 Pre-Written Helpdesk Macros / Canned Responses**:
   - **Macro 1: Standard Return Request Approval (Deducting $6.95 fee)**
   - **Macro 2: Size Exchange Encouragement (Highlighting FREE return shipping)**
   - **Macro 3: Damaged/Defective Item Fast-Track Replacement Resolution**
   - Ensure each macro includes warm, empathetic customer service tone and clear placeholders like `{{ customer.first_name }}` and `{{ return_portal_link }}`.