Food Finder

Food Finder

Recommend nearby restaurants based on the user's current location, time of day, and weather conditions.

Steps

1. Get Context

Call clawpaw_get_context to get:

• Current time and time period (morning/noon/afternoon/evening/late night)
• Current location coordinates

2. Get Weather

Call clawpaw_weather (using current coordinates, forecast=false) for real-time weather. Note the weather condition and temperature for recommendation adjustments.

3. Choose Food Category

Select search keywords based on time of day:

If the user mentions a specific cuisine (e.g. "hotpot", "coffee"), prioritize their preference over time-based defaults.

4. Search Nearby

Call clawpaw_nearby with:

• longitude / latitude: from context
• keywords: based on time period (or user preference)
• poi_type: 餐饮
• radius: adjust by weather
  • Rain/snow/extreme weather → 500
  • Normal weather → 1000
  • User has transport → 2000
• limit: 10

5. Output Recommendations

Pick the TOP 5 results and format:

🍜 Based on your location and the time of day, here are some picks:

1. **Restaurant Name** — XXm away
   ⭐ X.X | 💰 ¥XX avg | 📞 phone
   💡 [Reason tied to weather/time, e.g. "A warm bowl of soup on a rainy day"]

2. ...

Recommendation reasons should feel contextual:

• Rainy → "Close by, just X min walk"
• Cold → "Hot food to warm you up"
• Hot → "Air-conditioned, great for cooling off"
• Late night → "Open late, perfect for a midnight snack"

When to Use

• User expresses hunger ("I'm hungry", "starving")
• User asks what to eat ("what should I eat", "any food nearby")
• User asks about nearby restaurants ("any good restaurants around here")
• User mentions a specific food type ("I want hotpot", "coffee shop nearby")

When NOT to Use

• User is asking for a recipe or cooking instructions
• User specifies a different address (not their current location)
• User is discussing nutrition or dietary health topics