Add Airbnb

community_usecase/Airbnb-MCP/Airbnb_MCP.py

@@ -0,0 +1,127 @@
+import asyncio
+import sys
+from pathlib import Path
+from typing import List, Dict
+
+from dotenv import load_dotenv
+from camel.models import ModelFactory
+from camel.toolkits import FunctionTool, MCPToolkit
+from camel.types import ModelPlatformType, ModelType
+from camel.logger import set_log_level
+
+from owl.utils.enhanced_role_playing import OwlRolePlaying, arun_society
+
+import pathlib
+
+set_log_level(level="DEBUG")
+
+# Load environment variables from .env file if available
+load_dotenv()
+
+
+async def construct_society(
+    question: str,
+    tools: List[FunctionTool],
+) -> OwlRolePlaying:
+    """Build a multi-agent OwlRolePlaying instance with enhanced content curation capabilities."""
+    models = {
+        "user": ModelFactory.create(
+            model_platform=ModelPlatformType.OPENAI,
+            model_type=ModelType.GPT_4O,
+            model_config_dict={
+                "temperature": 0.7,
+                # "max_tokens": 4000  # Add token limit to prevent overflow
+            },
+        ),
+        "assistant": ModelFactory.create(
+            model_platform=ModelPlatformType.OPENAI,
+            model_type=ModelType.GPT_4O,
+            model_config_dict={
+                "temperature": 0.7,
+                #"max_tokens": 4000
+            },
+        ),
+    }
+
+    user_agent_kwargs = {"model": models["user"]}
+    assistant_agent_kwargs = {
+        "model": models["assistant"],
+        "tools": tools,
+    }
+
+    task_kwargs = {
+        "task_prompt": question,
+        "with_task_specify": False,
+    }
+
+    return OwlRolePlaying(
+        **task_kwargs,
+        user_role_name="content_curator",
+        user_agent_kwargs=user_agent_kwargs,
+        assistant_role_name="research_assistant",
+        assistant_agent_kwargs=assistant_agent_kwargs,
+    )
+
+async def main():
+    config_path = Path(__file__).parent / "mcp_servers_config.json"
+    mcp_toolkit = MCPToolkit(config_path=str(config_path))
+
+    try:
+        await mcp_toolkit.connect()
+
+        default_task = (
+            "Find me the best Airbnb in Gurugram with a check-in date of 2025-06-01 "
+            "and a check-out date of 2025-06-07 for 2 adults. Return the top 5 listings with their names, "
+            "prices, and locations."
+        )
+
+        task = sys.argv[1] if len(sys.argv) > 1 else default_task
+
+        # Connect to all MCP toolkits
+        tools = [*mcp_toolkit.get_tools()]
+        society = await construct_society(task, tools)
+        
+        try:
+            # Add error handling for the society execution
+            result = await arun_society(society)
+            
+            # Handle the result properly
+            if isinstance(result, tuple) and len(result) == 3:
+                answer, chat_history, token_count = result
+            else:
+                answer = str(result)
+                chat_history = []
+                token_count = 0
+
+        except Exception as e:
+            print(f"Error during society execution: {str(e)}")
+            raise
+
+    finally:
+        # Cleanup
+        await asyncio.sleep(1)
+        tasks = [t for t in asyncio.all_tasks() if t is not asyncio.current_task()]
+        for task in tasks:
+            task.cancel()
+            try:
+                await task
+            except asyncio.CancelledError:
+                pass
+        
+        try:
+            await mcp_toolkit.disconnect()
+        except Exception as e:
+            print(f"Cleanup error (can be ignored): {e}")
+
+if __name__ == "__main__":
+    try:
+        asyncio.run(main())
+    except KeyboardInterrupt:
+        print("\nShutting down gracefully...")
+    finally:
+        if sys.platform == 'win32':
+            try:
+                import asyncio.windows_events
+                asyncio.windows_events._overlapped = None
+            except (ImportError, AttributeError):
+                pass

community_usecase/Airbnb-MCP/README.md

@@ -0,0 +1,103 @@
+
+# 🏡 CAMEL-AI + MCP: Airbnb Use Case
+
+This example demonstrates how to use the [CAMEL-AI OWL framework](https://github.com/camel-ai/owl) and **MCP (Model Context Protocol)** to search for Airbnb listings using a custom MCP server (`@openbnb/mcp-server-airbnb`). Agents in the OWL framework coordinate to perform tool-augmented travel research in a structured, automated way.
+
+---
+
+## ✨ Use Case
+
+> _“Find me the best Airbnb in Gurugram with a check-in date of 2025-06-01 and a check-out date of 2025-06-07 for 2 adults. Return the top 5 listings with their names, prices, and locations.”_
+
+Agents leverage an MCP server to execute real-time Airbnb queries and return formatted results.
+
+---
+
+## 📦 Setup
+
+### 1. Clone and install dependencies
+
+```bash
+git clone https://github.com/camel-ai/owl
+cd owl
+pip install -r requirements.txt
+```
+
+---
+
+### 2. Configure MCP Server
+
+In your `mcp_servers_config.json`, add the following:
+
+```json
+{
+  "mcpServers": {
+    "airbnb": {
+      "command": "npx",
+      "args": [
+        "-y",
+        "@openbnb/mcp-server-airbnb",
+        "--ignore-robots-txt"
+      ]
+    }
+  }
+}
+```
+
+> 🛠️ You will need **Node.js and NPM** installed. Run `npx` will automatically fetch the Airbnb MCP server.
+
+---
+
+### 3. Run the Example Script
+
+```bash
+python community_usecase/Airbnb_MCP
+```
+
+You can also customize the prompt inside the script itself. Edit the `default_task` section of `Airbnb_MCP.py` like this:
+
+```python
+# Replace this line:
+default_task = (
+    "here you need to add the task"
+)
+
+# Example:
+default_task = (
+    "Find me the best Airbnb in Gurugram with a check-in date of 2025-06-01 "
+    "and a check-out date of 2025-06-07 for 2 adults. Return the top 5 listings with their names, "
+    "prices, and locations."
+)
+```
+
+This allows agents to work from your hardcoded task without passing anything via command line.
+
+---
+
+## 🧠 How It Works
+
+- **MCPToolkit** reads the config and connects to the `@openbnb/mcp-server-airbnb`.
+- **OWL RolePlaying Agents** simulate a conversation between a `content_curator` and a `research_assistant`.
+- The **assistant agent** calls the MCP Airbnb server to fetch listings.
+- The results are processed, formatted, and printed.
+
+---
+
+---
+
+## 🚧 Notes
+
+- This script uses **GPT-4o** via OpenAI for both user and assistant roles.
+- Supports async execution and graceful cleanup of agents and MCP sessions.
+- Add retries and fallback logic for production use.
+
+---
+
+## 📌 References
+
+- [MCP Overview (Anthropic)](https://docs.anthropic.com/en/docs/agents-and-tools/mcp)
+- [CAMEL-AI GitHub](https://github.com/camel-ai/camel)
+- [OWL Framework](https://github.com/camel-ai/owl)
+- [MCP Airbnb Plugin](https://www.npmjs.com/package/@openbnb/mcp-server-airbnb)
+~~~
+

community_usecase/Airbnb-MCP/mcp_servers_config.json

@@ -0,0 +1,12 @@
+{
+  "mcpServers": {
+    "airbnb": {
+      "command": "npx",
+      "args": [
+        "-y",
+        "@openbnb/mcp-server-airbnb",
+        "--ignore-robots-txt"
+      ]
+    }
+  }
+}