rappi-suggest

# Rappi Smart Suggestions Help the user decide what to eat based on their personal taste profile and ordering habits. ## Steps 1. Call `get_recommendations` with any context the user provided (e.g., "lunch", "something cheap", $ARGUMENTS). 2. Present recommendations grouped by type: **Your Usual** (type="usual"): - Show the store name and items they always order - "You order [items] from [store] regularly — want the usual?" - Action: offer to `quick_reorder` if order exists, or `add_to_cart` for each item **Right Now** (type="time_based"): - "You usually order from [store] around this time" - Action: offer to browse their menu with `get_restaurant_menu` **Try Something New** (type="new_store"): - "You haven't tried [store] yet" - Action: offer to `search_restaurants` or `get_restaurant_menu` 3. For each recommendation, offer a clear next action the user can take. 4. If the user wants deeper insight into their habits, call `get_taste_profile` and present: - Top food categories with percentages ("Hamburguesas 40%, Bebidas 25%") - Favorite stores and how often they order from each - Spending patterns (average per order, total this month) - Time habits ("You mostly order at lunch") - Top topping preferences 5. Additional context available: - `explore_verticals()` — show all available store types if they want to try something different - `get_credits_balance()` — mention credits balance if relevant to budget - `get_rappi_favorites()` — their favorited stores from the Rappi app - `browse_stores(store_type)` — browse specific store types (Turbo, markets, pharmacies) 6. If the user picks a recommendation, transition into the ordering flow: - Use `quick_reorder` for "usual" recommendations - Use `get_restaurant_menu` + `add_to_cart` for specific stores - For Turbo/market stores, use `search_store_products(store_id, query)` + `add_to_cart(... product_name, product_price)` - Always confirm before checkout 7. Prices in COP, formatted as $35.500.