product-search

name	product-search
description	Strategies for effective product catalog search and discovery
version	1.0
authors	["Commerce Team"]
tags	["search","discovery","products","catalog"]

Product Search Skill

When to Use

• User asks to find products by name, category, or feature
• User describes what they need without knowing exact product
• Need to discover products matching criteria
• Building a shortlist from the catalog

Search Strategies

By Keyword

Search using terms from:

• Product names (e.g., "headphones", "yoga mat")
• Product tags (e.g., "wireless", "eco-friendly")
• Features (e.g., "noise cancelling", "waterproof")
• Categories (e.g., "Electronics", "Apparel")

By Category

Available categories:

• Electronics: Audio, Wearables, Accessories
• Apparel: Tops, Footwear
• Home & Kitchen: Drinkware
• Sports & Outdoors: Fitness

By Price Range

Filter results by budget:

• Budget: Under $50
• Mid-range: $50-$150
• Premium: Over $150

Search Flow

1. IDENTIFY user intent
   └─ What are they looking for?
   └─ Any constraints (price, category, features)?

2. USE search_products tool
   └─ Start broad, then narrow
   └─ Try multiple keyword variations

3. FILTER results
   └─ Apply price constraints
   └─ Match required features

4. PRESENT findings
   └─ Top 3-5 matches
   └─ Include key differentiators
   └─ Note price and rating

Best Practices

Do

• Start with broad searches
• Try synonyms (e.g., "speaker" and "audio")
• Use category filters to narrow results
• Consider product tags for feature matching

Don't

• Return too many options (3-5 is ideal)
• Ignore stated constraints
• Skip rating/review information
• Forget to check availability

Example Searches

User Request	Search Strategy
"Something to listen to music"	search "audio" or "wireless"
"Gift under $50"	search all, filter by price
"Running gear"	search "running" + filter Apparel
"Tech accessories"	filter Electronics/Accessories

Quality Signals

When presenting results, highlight:

• Rating: 4.5+ is excellent, 4.0+ is good
• Reviews: More reviews = more reliable rating
• Price-to-value: Compare features to price point
• Brand: Match to user's preferred brands if known