Mapbox Search Patterns Skill - Expert guidance on choosing search tools and parameters for geocoding and POI search

mapbox-search-patterns-SKILL.md

name	description
mapbox-search-patterns	Expert guidance on choosing the right Mapbox search tool and parameters for geocoding, POI search, and location discovery

Mapbox Search Patterns Skill

Expert guidance for AI assistants on using Mapbox search tools effectively. Covers tool selection, parameter optimization, and best practices for geocoding, POI search, and location discovery.

Available Search Tools

1. search_and_geocode_tool

Best for: Specific places, addresses, brands, named locations

Use when query contains:

• Specific names: "Starbucks on 5th Avenue", "Empire State Building"
• Brand names: "McDonald's", "Whole Foods"
• Addresses: "123 Main Street, Seattle", "Times Square"
• Chain stores: "Target near me"
• Cities/places: "San Francisco", "Central Park"

Don't use for: Generic categories ("coffee shops", "museums")

2. category_search_tool

Best for: Generic place types, categories, plural queries

Use when query contains:

• Generic types: "coffee shops", "restaurants", "gas stations"
• Plural forms: "museums", "hotels", "parks"
• Is-a phrases: "any coffee shop", "all restaurants", "nearby pharmacies"
• Industry terms: "electric vehicle chargers", "ATMs"

Don't use for: Specific names or brands

3. reverse_geocode_tool

Best for: Converting coordinates to addresses

Use when:

• Have GPS coordinates, need human-readable address
• Need to identify what's at a specific location
• Converting user location to address

Tool Selection Decision Matrix

User Query	Tool	Reasoning
"Find Starbucks on Main Street"	search_and_geocode_tool	Specific brand name
"Find coffee shops nearby"	category_search_tool	Generic category, plural
"What's at 37.7749, -122.4194?"	reverse_geocode_tool	Coordinates to address
"Empire State Building"	search_and_geocode_tool	Specific named place
"hotels in downtown Seattle"	category_search_tool	Generic type + location
"Target store locations"	search_and_geocode_tool	Brand name (even plural)
"any restaurant near me"	category_search_tool	Generic + "any" phrase
"123 Main St, Boston, MA"	search_and_geocode_tool	Specific address
"electric vehicle chargers"	category_search_tool	Industry category
"McDonald's"	search_and_geocode_tool	Brand name

Parameter Guidance

Proximity vs Bbox vs Country

Three ways to spatially constrain search results:

1. proximity (STRONGLY RECOMMENDED)

What it does: Biases results toward a location, but doesn't exclude distant matches

Use when:

• User says "near me", "nearby", "close to"
• Have a reference point but want some flexibility
• Want results sorted by relevance to a point

Example:

{
  "q": "pizza",
  "proximity": {
    "longitude": -122.4194,
    "latitude": 37.7749
  }
}

Why this works: API returns SF pizza places first, but might include famous NYC pizzerias if highly relevant

⚠️ Critical: Always set proximity when you have a reference location! Without it, results are IP-based or global.

2. bbox (Bounding Box)

What it does: Hard constraint - ONLY returns results within the box

Use when:

• User specifies an area: "in downtown", "within this neighborhood"
• Have a defined service area
• Need to guarantee results are within bounds

Example:

{
  "q": "hotel",
  "bbox": [-122.51, 37.70, -122.35, 37.83]  // [minLon, minLat, maxLon, maxLat]
}

Why this works: Guarantees all hotels are within SF's downtown area

⚠️ Watch out: Too small = no results; too large = irrelevant results

3. country

What it does: Limits results to specific countries

Use when:

• User specifies country: "restaurants in France"
• Building country-specific features
• Need to respect regional boundaries

Example:

{
  "q": "Paris",
  "country": ["FR"]  // ISO 3166 alpha-2 codes
}

Why this works: Finds Paris, France (not Paris, Texas)

Can combine: proximity + country = "Near SF but only in USA"

Decision Matrix: Spatial Filters

Scenario	Use	Why
"Find coffee near me"	proximity	Bias toward user location
"Coffee shops in downtown Seattle"	proximity + bbox	Center on downtown, limit to area
"Hotels in France"	country	Hard country boundary
"Best pizza in San Francisco"	proximity + country ["US"]	Bias to SF, limit to US
"Gas stations along this route"	bbox around route	Hard constraint to route corridor
"Restaurants within 5 miles"	proximity (then filter by distance)	Bias nearby, filter results

Setting limit Parameter

category_search_tool only (1-25, default 10)

Use Case	Limit	Reasoning
Quick suggestions	5	Fast, focused results
Standard list	10	Default, good balance
Comprehensive search	25	Maximum allowed
Map visualization	25	Show all nearby options
Dropdown/autocomplete	5	Don't overwhelm UI

Performance tip: Lower limits = faster responses

types Parameter (search_and_geocode_tool)

Filter by feature type:

Type	What It Includes	Use When
poi	Points of interest (businesses, landmarks)	Looking for places, not addresses
address	Street addresses	Need specific address
place	Cities, neighborhoods, regions	Looking for area/region
street	Street names without numbers	Need street, not specific address
postcode	Postal codes	Searching by ZIP/postal code
district	Districts, neighborhoods	Area-based search
locality	Cities, towns, villages	Municipality search
country	Country names	Country-level search

Example combinations:

// Only POIs and addresses, no cities
{"q": "Paris", "types": ["poi", "address"]}
// Returns Paris Hotel, Paris Street, not Paris, France

// Only places (cities/regions)
{"q": "Paris", "types": ["place"]}
// Returns Paris, France; Paris, Texas; etc.

Default behavior: All types included (usually what you want)

poi_category Parameter

search_and_geocode_tool: Narrow generic searches

{
  "q": "lunch",
  "poi_category": ["restaurant", "cafe"],
  "proximity": {"longitude": -122.4194, "latitude": 37.7749}
}

When to use:

• Generic query that could match multiple categories
• Want to focus search within category
• User specifies type implicitly

category_search_tool: Use poi_category_exclusions instead

{
  "category": "food_and_drink",
  "poi_category_exclusions": ["bar", "nightclub"]
}

When to use:

• Broad category but want to exclude subcategories
• "Restaurants but not fast food"

auto_complete Parameter (search_and_geocode_tool)

What it does: Enables partial/fuzzy matching

Setting	Behavior	Use When
true	Matches partial words, typos	User typing in real-time
false (default)	Exact matching	Final query, not autocomplete

Example:

// User types "starb"
{"q": "starb", "auto_complete": true}
// Returns: Starbucks, Starboard Tavern, etc.

Use for:

• Search-as-you-type interfaces
• Handling typos ("mcdonalds" → McDonald's)
• Incomplete queries

Don't use for:

• Final/submitted queries (less precise)
• When you need exact matches

ETA Parameters (search_and_geocode_tool)

Request estimated time of arrival to results

Parameters:

• eta_type: Set to "navigation"
• navigation_profile: "driving" | "walking" | "cycling"
• origin: Starting coordinates

Use when:

• User asks "how long to get there?"
• Sorting by travel time, not distance
• Need route time, not straight-line distance

Example:

{
  "q": "grocery stores",
  "proximity": {"longitude": -122.4194, "latitude": 37.7749},
  "eta_type": "navigation",
  "navigation_profile": "driving",
  "origin": {"longitude": -122.4194, "latitude": 37.7749}
}

Returns: Results with eta (travel time in seconds)

⚠️ Cost: Requires routing calculation per result (counts toward API quota)

When NOT to use:

• Just need straight-line distance (use distance_tool offline after search)
• Budget-conscious (adds API cost)

format Parameter (category_search_tool)

Choose output format:

Format	Returns	Use When
formatted_text (default)	Human-readable text	Displaying to user directly
json_string	GeoJSON as JSON string	Need to parse/process results

Example:

formatted_text:

1. Blue Bottle Coffee
   Address: 66 Mint St, San Francisco, CA
   Coordinates: 37.7825, -122.4052
   Type: poi

json_string:

{
  "type": "FeatureCollection",
  "features": [{
    "type": "Feature",
    "geometry": {"type": "Point", "coordinates": [-122.4052, 37.7825]},
    "properties": {"name": "Blue Bottle Coffee", ...}
  }]
}

Decision:

• Showing list to user → formatted_text
• Plotting on map → json_string (parse and use coordinates)
• Further processing → json_string

language Parameter

ISO language codes (e.g., "en", "es", "fr", "de", "ja", "zh")

Use when:

• Building multilingual app
• User's language preference known
• Need localized names

Example:

{
  "q": "東京タワー",
  "language": "ja"
}
// Returns results in Japanese

Default: English (if not specified)

Tip: Match user's locale for best experience

Common Patterns and Workflows

Pattern 1: "Near Me" Search

User: "Find coffee shops near me"

Optimal approach:

1. Get user's location (from app/browser)
2. Use category_search_tool:
   - category: "cafe" or "coffee_shop"
   - proximity: user's coordinates
   - limit: 10

Why: Category tool for generic "coffee shops", proximity for "near me"

Pattern 2: Branded Chain Lookup

User: "Find all Starbucks in Seattle"

Optimal approach:

1. Use search_and_geocode_tool:
   - q: "Starbucks"
   - proximity: Seattle coordinates
   - country: ["US"]
2. Or if need strict boundary:
   - bbox: Seattle city bounds

Why: Brand name = search_and_geocode_tool; proximity biases to Seattle

Pattern 3: Address Geocoding

User: "What are the coordinates of 1600 Pennsylvania Ave?"

Optimal approach:

Use search_and_geocode_tool:
- q: "1600 Pennsylvania Ave, Washington DC"
- types: ["address"]  // Focus on addresses
- country: ["US"]     // Narrow to US

Why: Specific address with country context for disambiguation

Pattern 4: Category Search with Area Restriction

User: "Show me all hotels in downtown Portland"

Optimal approach:

1. Geocode "downtown Portland" → get center point
2. Define downtown bbox (or use 1-2 mile radius)
3. Use category_search_tool:
   - category: "hotel"
   - bbox: downtown bounds (or proximity + filter by distance)
   - limit: 25  // Get comprehensive list

Why: Category for "hotels", bbox for "in downtown" hard boundary

Pattern 5: Reverse Geocoding

User: "What's at these GPS coordinates?"

Optimal approach:

Use reverse_geocode_tool:
- longitude: -122.4194
- latitude: 37.7749
- types: ["address", "poi"]  // Get both address and POI if available

Why: Coordinates → address is exactly what reverse geocoding does

Pattern 6: Route-Based Search

User: "Find gas stations along my route"

Optimal approach:

1. Get route geometry from directions_tool
2. Create bbox around route (use bounding_box_tool)
3. Use category_search_tool:
   - category: "gas_station"
   - bbox: route bounding box
4. Filter results to those within X meters of route (use distance_tool)

Why: Bbox for rough filter, then distance calculation for precision

Pattern 7: Multilingual POI Search

User: "Find ramen shops" (user locale: ja)

Optimal approach:

Use category_search_tool:
- category: "ramen_restaurant" (or "restaurant")
- language: "ja"
- proximity: user location

Why: Returns Japanese names/addresses for better UX

Anti-Patterns to Avoid

❌ Don't: Use category_search for brands

// BAD
category_search_tool({category: "starbucks"})
// "starbucks" is not a category, returns error

// GOOD
search_and_geocode_tool({q: "Starbucks"})

❌ Don't: Use search_and_geocode for generic categories

// BAD
search_and_geocode_tool({q: "coffee shops"})
// Less precise, may return unrelated results

// GOOD
category_search_tool({category: "cafe"})

❌ Don't: Forget proximity for local searches

// BAD - Results may be anywhere globally
category_search_tool({category: "restaurant"})

// GOOD - Biased to user location
category_search_tool({
  category: "restaurant",
  proximity: {longitude: -122.4194, latitude: 37.7749}
})

❌ Don't: Use bbox when you mean proximity

// BAD - Hard boundary may exclude good nearby results
search_and_geocode_tool({
  q: "pizza",
  bbox: [-122.42, 37.77, -122.41, 37.78]  // Tiny box
})

// GOOD - Bias toward point, but flexible
search_and_geocode_tool({
  q: "pizza",
  proximity: {longitude: -122.4194, latitude: 37.7749}
})

❌ Don't: Request ETA unnecessarily

// BAD - Costs API quota for routing calculations
search_and_geocode_tool({
  q: "museums",
  eta_type: "navigation",
  navigation_profile: "driving"
})
// User didn't ask for travel time!

// GOOD - Only add ETA when needed
search_and_geocode_tool({q: "museums"})
// If user asks "how long to get there?", then add ETA

❌ Don't: Set limit too high for UI display

// BAD - Overwhelming for simple dropdown
category_search_tool({
  category: "restaurant",
  limit: 25
})
// Returns 25 restaurants for a 5-item dropdown

// GOOD - Match UI needs
category_search_tool({
  category: "restaurant",
  limit: 5
})

Performance Optimization

Minimize API Calls

Pattern: Geocode once, reuse coordinates

// GOOD
1. User enters "Seattle"
2. Geocode "Seattle" → (lng, lat)
3. Use those coordinates for multiple category searches
4. Cache coordinates for session

// BAD
1. Geocode "Seattle" for coffee search
2. Geocode "Seattle" again for restaurant search
3. Geocode "Seattle" again for hotel search

Set Appropriate Limits

UI Context	Recommended Limit
Autocomplete dropdown	5
List view	10
Map view	25
Export/download	25 (or paginate)

Use Offline Tools When Possible

After getting search results:

1. category_search_tool → Get POIs
2. distance_tool (offline) → Calculate distances
3. bearing_tool (offline) → Get directions

Why: Search once (API), then use offline tools for calculations (free, fast)

Combining Search with Other Tools

Search → Distance Calculation

1. category_search_tool({category: "hospital", proximity: user_location})
   → Returns 10 hospitals with coordinates
2. distance_tool(user_location, each_hospital)
   → Calculate exact distances offline
3. Sort by distance

Search → Directions

1. search_and_geocode_tool({q: "Space Needle"})
   → Get destination coordinates
2. directions_tool({from: user_location, to: space_needle_coords})
   → Get turn-by-turn directions

Search → Isochrone → Containment Check

1. search_and_geocode_tool({q: "warehouse"})
   → Get warehouse coordinates
2. isochrone_tool({coordinates: warehouse, time: 30, profile: "driving"})
   → Get 30-minute delivery zone polygon
3. point_in_polygon_tool(customer_address, delivery_zone)
   → Check if customer is in delivery zone

Search → Static Map Visualization

1. category_search_tool({category: "restaurant", limit: 10})
   → Get restaurant coordinates
2. static_map_image_tool({
     markers: restaurant_coordinates,
     auto_fit: true
   })
   → Create map image showing all restaurants

Handling No Results

If category_search returns no results:

Possible reasons:

1. Invalid category → Use resource_reader_tool with mapbox://categories to see valid categories
2. Too restrictive bbox → Expand area or use proximity instead
3. No POIs in area → Try broader category or remove spatial filters
4. Wrong country filter → Check country codes

Example recovery:

1. category_search_tool({category: "taco"}) → No results
2. Check: Is "taco" a valid category?
   → Use category_list_tool → See "mexican_restaurant" is valid
3. Retry: category_search_tool({category: "mexican_restaurant"}) → Success

If search_and_geocode returns no results:

Possible reasons:

1. Typo in query → Retry with auto_complete: true
2. Too specific → Broaden search (remove address numbers, try nearby city)
3. Wrong types filter → Remove or expand types
4. Not a recognized place → Check spelling, try alternative names

Category List Resource

Get valid categories: Use resource_reader_tool or category_list_tool

resource_reader_tool({uri: "mapbox://categories"})

Returns: All valid category IDs (e.g., "restaurant", "hotel", "gas_station")

When to use:

• User enters free-text category
• Need to map user terms to Mapbox categories
• Validating category before search

Example mapping:

• User: "places to eat" → Category: "restaurant"
• User: "gas" → Category: "gas_station"
• User: "lodging" → Category: "hotel"

Quick Reference

Tool Selection Flowchart

User query contains...

→ Specific name/brand (Starbucks, Empire State Building)
  → search_and_geocode_tool

→ Generic category/plural (coffee shops, museums, any restaurant)
  → category_search_tool

→ Coordinates → Address
  → reverse_geocode_tool

→ Address → Coordinates
  → search_and_geocode_tool with types: ["address"]

Essential Parameters Checklist

For local searches, ALWAYS set:

• ✅ proximity (or bbox if strict boundary needed)

For category searches, consider:

• ✅ limit (match UI needs)
• ✅ format (json_string if plotting on map)

For disambiguation, use:

• ✅ country (when geographic context matters)
• ✅ types (when feature type matters)

For travel-time ranking:

• ✅ eta_type, navigation_profile, origin (costs API quota)

Common Mistakes

1. Forgetting proximity → Results are global/IP-based
2. Using wrong tool → category_search for "Starbucks" (use search_and_geocode)
3. Invalid category → Check category_list first
4. Bbox too small → No results; use proximity instead
5. Requesting ETA unnecessarily → Adds API cost
6. Limit too high for UI → Overwhelming user
7. Not filtering types → Get cities when you want POIs

Integration with Other Skills

Works with:

• mapbox-geospatial-operations: After search, use offline distance/bearing calculations
• mapbox-web-integration-patterns: Display search results on map in web app
• mapbox-token-security: Ensure search requests use properly scoped tokens

Resources

• Mapbox Search Box API Docs
• Category Search API
• Geocoding API
• Category List Resource