Lamatic · sujalchettri01 · Apr 16, 2026 · May 19, 2026 · May 19, 2026 · May 19, 2026
diff --git a/kits/smart-travel-planner/Readme.md b/kits/smart-travel-planner/Readme.md
@@ -0,0 +1,322 @@
+#  Smart Travel Planner
+
+## Live Demo
+
+https://ai-travel-indol.vercel.app/
+
+> An **Agentic AI-powered** travel planning platform built with **Next.js** and **Lamatic AI** that generates complete, structured travel itineraries with maps, photos, local food, cultural highlights, and an interactive chatbot.
+
+---
+
+##  Features
+
+-  **AI-Generated Itineraries** — Day-wise structured travel plans powered by Lamatic AI LLM flows
+-  **Live Maps** — OpenStreetMap embedded per day via Geocode API
+-  **Destination & Day Photos** — Auto-fetched via Places API with lightbox viewer
+-  **Local Food & Stay Suggestions** — Curated per day and per destination
+-  **Cultural Highlights & Travel Tips** — Structured output from the LLM
+-  **Floating Travel Chatbot** — Ask follow-up travel questions in real time
+-  **Fallback Flow** — Condition-based routing ensures users always get a response
+-  **GraphQL Workflow Execution** — Lamatic AI called via GraphQL `executeWorkflow` query
+
+---
+
+##  Tech Stack
+
+| Layer | Technology |
+|-------|------------|
+| Frontend | Next.js 14+ (App Router, TypeScript) |
+| Styling | CSS Modules / globals.css |
+| AI Orchestration | Lamatic AI (Agentic Workflows via GraphQL) |
+| Maps | OpenStreetMap (iframe embed) |
+| Location | Geocode API (lat/lng resolution) |
+| Photos | Google Places API |
+| Language | TypeScript (TSX) |
+| Runtime | Node.js |
+
+---
+
+##  Project Structure
+
+```
+smart-travel-planner/
+├── apps/
+│   ├── actions/
+│   │   └── orchestrate.ts         # Lamatic AI workflow runners
+│   ├── app/
+│   │   ├── api/
+│   │   │   ├── chatbot/route.ts   # POST /api/chatbot
+│   │   │   ├── geocode/route.ts   # GET /api/geocode
+│   │   │   ├── places/route.ts    # GET /api/places
+│   │   │   └── travel/route.ts    # POST /api/travel
+│   │   ├── globals.css
+│   │   ├── layout.tsx
+│   │   └── page.tsx
+│   ├── components/
+│   │   ├── DayCard.tsx
+│   │   ├── FloatingChatbot.tsx
+│   │   ├── ItineraryDisplay.tsx
+│   │   └── PlannerForm.tsx
+│   ├── lib/
+│   │   ├── lamatic-client.ts      # Lamatic SDK client
+│   │   └── utils.ts
+│   ├── public/
+│   │   └── logo.png
+│   ├── .env.example
+│   ├── next.config.ts
+│   ├── package.json
+│   └── tsconfig.json
+├── constitutions/
+│   └── default.md                 # Agent guardrails
+├── flows/
+│   ├── chatbot-flow.ts            # Chatbot Lamatic flow
+│   └── travel-flow.ts             # Travel itinerary Lamatic flow
+├── agent.md                       # Agent identity doc
+├── lamatic.config.ts              # Kit metadata
+└── Readme.md
+```
+---
+
+##  Lamatic AI Agent Flows
+
+###  Travel Itinerary Flow
+
+```
+API Request  →  Code Node  →  LLM (Generate Itinerary)
+                                        │
+                               Condition Node (validate output)
+                                ┌───────┴────────┐
+                            Valid             Invalid
+                         Refine Text        Fallback Text
+                                └───────┬────────┘
+                                   code2 Node (parse JSON)
+                                        │
+                                   API Response
+```
+
+| Node | Purpose |
+|------|---------|
+| **API Request** | Receives `destination`, `days`, `budget`, `destination_type` |
+| **Code Node** | Pre-processes and formats input for the LLM |
+| **Generate Itinerary (LLM)** | Core AI node — generates full structured travel plan |
+| **Condition Node** | Validates output completeness and structure |
+| **Refine Generate Text** | Cleans valid output into structured JSON |
+| **Fallback Generate Text** | Generates a simpler plan if the main LLM fails |
+| **code2 Node** | Final JSON parse and validation |
+| **API Response** | Returns structured itinerary to the frontend |
+
+###  Chatbot Flow
+
+```
+API Request (user message)  →  Generate Text (LLM)  →  API Response
+```
+
+The chatbot is scoped to travel-only questions: destinations, food, budget, culture, visas, and travel tips.
+
+---
+
+##  Lamatic Workflow Integration
+
+All Lamatic AI flows are called through a shared GraphQL utility in `actions/orchestrate.ts`:
+
+```typescript
+// Generic workflow runner
+export async function runLamaticWorkflow({ workflowId, payload }) {
+  const res = await fetch(process.env.LAMATIC_API_URL, {
+    method: "POST",
+    headers: {
+      "Content-Type": "application/json",
+      Authorization: `Bearer ${process.env.LAMATIC_API_KEY}`,
+      "x-project-id": process.env.LAMATIC_PROJECT_ID,
+    },
+    body: JSON.stringify({
+      query: `
+        query Execute($workflowId: String!, $payload: JSON!) {
+          executeWorkflow(workflowId: $workflowId, payload: $payload) {
+            status
+            result
+          }
+        }
+      `,
+      variables: { workflowId, payload },
+    }),
+    cache: "no-store",
+  });
+}
+
+// Travel agent — uses GLOBAL_TRAVEL_AGENT workflow
+export async function runTravelAgent(input: {
+  destination: string;
+  days: number;
+  budget: number;
+  destination_type: string;
+}) { ... }
+
+// Chatbot agent — uses CHATBOT_FLOW_ID workflow
+export async function runChatAgent({ message }: { message: string }) { ... }
+```
+
+---
+
+##  Key Component — `PlannerForm.tsx`
+
+The main component handles the complete UI lifecycle:
+
+**State managed:**
+```typescript
+destination, days, budget, destinationType   // Form inputs
+loading, response                             // Request state
+activeDay                                     // Selected day tab
+photos                                        // Destination photo gallery
+lightboxImg                                   // Lightbox fullscreen image
+dayLocation: { lat, lng }                     // Map coordinates per day
+dayPhotos                                     // Per-day photo gallery
+```
+
+**Data flow on submit:**
+```
+handleSubmit()
+  → POST /api/travel         (generate itinerary)
+  → GET /api/places          (fetch destination photos)
+  → fetchDayLocation()       (GET /api/geocode for Day 1 map)
+  → fetchDayPhoto()          (GET /api/places for Day 1 photos)
+```
+
+**On day tab switch:**
+```
+onClick (Day tab)
+  → fetchDayLocation(morning_location || afternoon_location || destination)
+  → fetchDayPhoto(location)
+```
+
+**Map rendering** uses OpenStreetMap via iframe with dynamic bbox:
+```typescript
+src={`https://www.openstreetmap.org/export/embed.html
+  ?bbox=${lng - 0.05},${lat - 0.05},${lng + 0.05},${lat + 0.05}
+  &layer=mapnik&marker=${lat},${lng}`}
+```
+
+---
+
+##  API Endpoints
+
+| Method | Endpoint | Description |
+|--------|----------|-------------|
+| `POST` | `/api/travel` | Generate full travel itinerary |
+| `POST` | `/api/chatbot` | Send message to travel chatbot |
+| `GET` | `/api/places?query=` | Fetch destination/place photos |
+| `GET` | `/api/geocode?query=` | Resolve place name to lat/lng |
+
+### Example — Generate Itinerary
+
+**Request:**
+```json
+POST /api/travel
+{
+  "destination": "Bali",
+  "days": 5,
+  "budget": 1000,
+  "destination_type": "beach"
+}
+```
+
+**Response shape:**
+```typescript
+{
+  success: boolean;
+  itinerary: {
+    destination: string;
+    country: string;
+    introduction: string;
+    best_time_to_visit: string;
+    estimated_budget: string;
+    highlights: { name: string; description: string }[];
+    food: { name: string; description: string }[];
+    culture: string[];
+    travel_tips: string[];
+    days: {
+      day: number;
+      title: string;
+      morning: string;
+      afternoon: string;
+      evening: string;
+      morning_location?: string;
+      afternoon_location?: string;
+      evening_location?: string;
+      food_recommendation: string;
+      stay_suggestion: string;
+      estimated_day_cost: string;
+      notes: string;
+    }[];
+  }
+}
+```
+
+---
+
+##  Getting Started
+
+### Prerequisites
+
+- Node.js v18+
+- npm or yarn
+- Lamatic AI account with configured workflows
+- Google Places API key
+- Geocoding API key
+
+### Installation
+
+```bash
+git clone https://github.com/your-username/smart-travel.git
+cd smart-travel
+npm install
+```
+
+### Environment Variables
+
+Create a `.env.local` file:
+
+```env
+# Lamatic AI
+LAMATIC_API_URL=https://your-lamatic-endpoint/graphql
+LAMATIC_API_KEY=your_lamatic_api_key
+LAMATIC_PROJECT_ID=your_project_id
+GLOBAL_TRAVEL_AGENT=your_travel_workflow_id
+CHATBOT_FLOW_ID=your_chatbot_workflow_id
+
+# APIs
+GOOGLE_PLACES_API_KEY=your_google_places_api_key
+GEOCODE_API_KEY=your_geocode_api_key
+```
+
+### Run Development Server
+
+```bash
+npm run dev
+```
+
+Open [http://localhost:3000](http://localhost:3000)
+
+---
+
+##  Built With Lamatic AI
+
+This project was built for the **Lamatic AI Agentic Kit Challenge**. It uses Lamatic's agentic flow infrastructure with multi-node pipelines, condition-based routing, structured JSON outputs, and fallback handling — all connected to a Next.js frontend.
+
+---
+
+##  Acknowledgements
+
+- [Lamatic AI](https://lamatic.ai) — Agentic workflow platform
+- [Next.js](https://nextjs.org) — Frontend framework
+- [OpenStreetMap](https://openstreetmap.org) — Embedded maps
+- Google Places & Geocode APIs — Photos and location resolution
+
+---
+
+##  License
+
+MIT License — feel free to use and adapt this project.
+
+---
+
diff --git a/kits/smart-travel-planner/agent.md b/kits/smart-travel-planner/agent.md
@@ -0,0 +1,29 @@
+# Smart AI Travel Planner Agent
+
+## Identity
+I am an agentic AI-powered travel planning assistant built on Lamatic AI workflows. I generate complete, structured travel itineraries with maps, photos, local food suggestions, cultural highlights, and real-time chatbot assistance.
+
+## Capabilities
+- Generate day-by-day structured travel itineraries for any destination
+- Provide live maps per day using OpenStreetMap and Geocode API
+- Fetch destination and day-wise photos via Google Places API
+- Suggest local food, stay options, and estimated costs per day
+- Share cultural highlights and travel tips for destinations
+- Answer follow-up travel questions via a floating chatbot
+- Handle fallback routing when the main LLM output is incomplete
+
+## Flows
+- **Travel Itinerary Flow** — Takes destination, days, budget, and destination type as input. Routes through a Code Node → LLM → Condition Node → Refine/Fallback → JSON parse → API Response
+- **Chatbot Flow** — Takes a user message and returns a travel-scoped AI response covering destinations, food, budget, culture, visas, and travel tips
+
+## Personality
+- Friendly, enthusiastic, and knowledgeable about travel
+- Clear and organized when presenting itinerary information
+- Culturally aware and sensitive to diverse travel needs
+- Helpful in suggesting alternatives when needed
+
+## Limitations
+- Cannot book flights, hotels, or activities directly
+- Real-time data depends on connected APIs (Google Places, Geocode)
+- Suggestions are AI-generated and should be verified before travel
+- Chatbot is scoped to travel-related questions only
diff --git a/kits/smart-travel-planner/apps/.env.example b/kits/smart-travel-planner/apps/.env.example
@@ -0,0 +1,9 @@
+LAMATIC_API_URL=your_lamatic_api_url
+LAMATIC_PROJECT_ID=your_project_id
+GLOBAL_TRAVEL_AGENT=your_agent_id
+LAMATIC_API_KEY=your_api_key
+LAMATIC_FLOW_URL=your_flow_url
+CHATBOT_FLOW_ID=your_chatbot_flow_id
+
+GOOGLE_PLACES_API_KEY=your_google_places_api_key
+NEXT_PUBLIC_GOOGLE_PLACES_API_KEY=your_google_places_api_key