adds support for Docker MCP Toolkit and creates a content search tool…

community_usecase/Docker-Toolkit-MCP/README.md

@@ -0,0 +1,147 @@
+# 🔍 CAMEL-AI + MCP: Docker Toolkit Search Tool
+
+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 Docker Toolkit related content. The tool leverages intelligent agents to gather and organize documentation, tutorials, best practices, and troubleshooting guides.
+
+---
+
+## ✨ Use Case
+
+> _"Search and organize Docker Toolkit related content including technical documentation, video tutorials, best practices, and optimization guides. Return results in a structured format with descriptions and links."_
+
+Agents coordinate to perform comprehensive searches and return well-organized results in Markdown format.
+
+---
+
+## 📦 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
+
+Create `mcp_servers_config.json` in the project directory:
+
+```json
+{
+  "mcpServers": {
+    "MCP_DOCKER": {
+      "type": "stdio",
+      "command": "docker",
+      "args": [
+        "run",
+        "-l",
+        "--rm",
+        "-i",
+        "alpine/socat",
+        "STDIO",
+        "TCP:host.docker.internal:8811"
+      ]
+    }
+  }
+}
+```
+
+> 🛠️ Make sure to set up your environment variables in `.env` file.
+
+### 3. Run the Search Tool
+
+```bash
+python docker_content_search.py
+```
+
+You can customize the search query by modifying the `default_task` in `docker_content_search.py`:
+
+```python
+default_task = (
+    "Please help me search for Docker Toolkit related content:\n"
+    "1. Docker Toolkit technical documentation and official guides\n"
+    "2. Docker Toolkit video tutorials and practical cases\n"
+    # ...customize your search requirements...
+)
+```
+
+---
+
+## 🧠 How It Works
+
+- **MCPToolkit** connects to the search service
+- **OWL RolePlaying Agents** simulate a conversation between:
+  - `docker_content_curator`: Manages content requirements
+  - `docker_tech_expert`: Executes searches and formats results
+- Results are saved as Markdown files in the `search_results` directory
+
+## 📄 Output Format
+
+Search results are organized in the following sections:
+- Documentation and Guides
+- Video Tutorials
+- Best Practices
+- Troubleshooting and Optimization
+
+Each entry includes:
+- Title with link
+- Brief description
+- Additional metadata (duration, author for videos)
+
+---
+
+## 🚧 Notes
+
+### Prerequisites
+- Docker Desktop must be installed and running
+- MCP Toolkit must be installed in Docker:
+- Firecrawl plugin must be configured in MCP Toolkit:
+  ![MCP Toolkit Firecrawl Configuration](docker-toolkit-firecrawl.png)
+
+### General Notes
+- Uses **DeepSeek** model for both user and assistant roles
+- Supports async execution and error handling
+- Includes retry mechanism for reliability
+- Saves results in timestamped Markdown files
+
+---
+
+## 🌟 Docker MCP Toolkit Advantages
+
+### Simplified Management
+- **One-Time Setup**: Configure Docker MCP once to access all MCP services in the Docker MCP marketplace
+- **Automated Resource Management**: 
+  - On-demand container startup
+  - Automatic credential injection
+  - Container termination after task completion
+  - Efficient resource utilization
+
+### Enhanced Security
+- **Runtime Protection**: 
+  - Container isolation protection
+  - Limited host system access
+  - Sandboxed environment
+- **Credential Management**:
+  - Built-in OAuth support
+  - Secure credential storage
+  - No plaintext environment variables
+
+### Developer Productivity
+- **Tool Discovery**: Verified high-quality MCP tools through Docker Hub
+- **Simplified Configuration**: 
+  - Automatic dependency handling
+  - Environment management
+  - No manual service hosting required
+- **Standardized Integration**: Unified interface and management approach
+
+Docker MCP Catalog and Toolkit's core design philosophy is to make MCP tools simple and secure to use, fundamentally changing how developers manage MCP tools by providing better structure, trustworthiness, and scalability.
+
+
+## 📌 References
+
+- [CAMEL-AI GitHub](https://github.com/camel-ai/camel)
+- [OWL Framework](https://github.com/camel-ai/owl)
+- [MCP Documentation](https://docs.anthropic.com/claude/docs/model-context-protocol)
+- [DeepSeek Models](https://api-docs.deepseek.com/)
+- [Docker Toolkit](https://www.docker.com/products/mcp-catalog-and-toolkit/)
+- [Docker MCP Toolkit Installation Guide](https://www.cloudesx.com/article/mcp-server-docker)

community_usecase/Docker-Toolkit-MCP/docker_content_search.py

@@ -0,0 +1,177 @@
+import asyncio
+import sys
+import datetime
+from pathlib import Path
+from typing import List
+
+from dotenv import load_dotenv
+from camel.models import ModelFactory
+from camel.toolkits import FunctionTool, MCPToolkit, SearchToolkit
+from camel.types import ModelPlatformType, ModelType
+from camel.logger import set_log_level
+
+from owl.utils.enhanced_role_playing import OwlRolePlaying, arun_society
+
+set_log_level(level="DEBUG")
+load_dotenv()
+
+async def construct_society(
+    question: str,
+    tools: List[FunctionTool],
+) -> OwlRolePlaying:
+    """Construct OwlRolePlaying instance to enhance content search capabilities"""
+    models = {
+        "user": ModelFactory.create(
+            model_platform=ModelPlatformType.DEEPSEEK,
+            model_type=ModelType.DEEPSEEK_CHAT,
+            model_config_dict={"temperature": 0.7},
+        ),
+        "assistant": ModelFactory.create(
+            model_platform=ModelPlatformType.DEEPSEEK,
+            model_type=ModelType.DEEPSEEK_CHAT,
+            model_config_dict={"temperature": 0.7},
+        ),
+    }
+
+    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="docker_content_curator",
+        user_agent_kwargs=user_agent_kwargs,
+        assistant_role_name="docker_tech_expert",
+        assistant_agent_kwargs=assistant_agent_kwargs,
+    )
+
+async def save_search_results(content: str, token_count: int = None) -> Path:
+    """Save search results as Markdown file
+
+    Args:
+        content: Search result content
+        token_count: Number of tokens used
+
+    Returns:
+        Path to saved file
+    """
+    # Create output directory
+    output_dir = Path(__file__).parent / "search_results"
+    output_dir.mkdir(exist_ok=True)
+
+    # Generate filename
+    timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
+    filename = f"docker_toolkit_search_{timestamp}.md"
+    output_path = output_dir / filename
+
+    # Organize file content
+    file_content = [
+        "# Docker Toolkit Search Results",
+        f"\n> Generated at: {datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')}",
+    ]
+
+    if token_count:
+        file_content.append(f"\n> Tokens used: {token_count}")
+
+    file_content.append("\n## Search Content\n")
+    file_content.append(content)
+
+    # Write to file
+    output_path.write_text("\n".join(file_content), encoding="utf-8")
+
+    return output_path
+
+async def main():
+    # Configure MCP toolkit and search toolkit
+    config_path = Path(__file__).parent / "mcp_servers_config.json"
+    mcp_toolkit = MCPToolkit(config_path=str(config_path))
+    try:
+        await mcp_toolkit.connect()
+
+        default_task = (
+            "Please help me search for Docker Toolkit related content:\n"
+            "1. Docker Toolkit technical documentation and official guides\n"
+            "2. Docker Toolkit video tutorials and practical cases\n"
+            "3. Docker Toolkit best practices in development and operations\n"
+            "4. Docker Toolkit performance optimization and troubleshooting guides\n"
+            "\n"
+            "Please note:\n"
+            "- Prioritize recent content\n"
+            "- Ensure content quality and reliability\n"
+            "- Include both English and Chinese resources\n"
+            "- Focus on practical experience and case studies\n"
+            "\n"
+            "Please organize in the following format:\n"
+            "## Documentation and Guides\n"
+            "- [Title](link)\n"
+            "  Brief description\n"
+            "\n"
+            "## Video Tutorials\n"
+            "- [Title](link)\n"
+            "  Duration | Author | Brief description\n"
+            "\n"
+            "## Best Practices\n"
+            "- [Title](link)\n"
+            "  Brief description\n"
+            "\n"
+            "## Troubleshooting and Optimization\n"
+            "- [Title](link)\n"
+            "  Brief description"
+        )
+
+        task = sys.argv[1] if len(sys.argv) > 1 else default_task
+
+        # Combine MCP tools and search tools
+        tools = [
+            *mcp_toolkit.get_tools(),
+        ]
+
+        society = await construct_society(task, tools)
+
+        try:
+            result = await arun_society(society)
+
+            if isinstance(result, tuple) and len(result) == 3:
+                answer, chat_history, token_count = result
+                print("\n🔍 Search Results:")
+                print("=" * 80)
+                print(answer)
+                print("=" * 80)
+                print(f"\n📊 Tokens used: {token_count}")
+
+                # Save search results
+                output_path = await save_search_results(answer, token_count)
+                print(f"\n💾 Results saved to: {output_path}")
+
+            else:
+                print("\n🔍 Search Results:")
+                print("=" * 80)
+                print(str(result))
+                print("=" * 80)
+
+                # Save search results
+                output_path = await save_search_results(str(result))
+                print(f"\n💾 Results saved to: {output_path}")
+
+        except Exception as e:
+            print(f"❌ Search execution error: {str(e)}")
+            raise
+
+    finally:
+        try:
+            await mcp_toolkit.disconnect()
+        except:
+            pass
+
+if __name__ == "__main__":
+    asyncio.run(main())

community_usecase/Docker-Toolkit-MCP/mcp_servers_config.json

@@ -0,0 +1,17 @@
+{
+  "mcpServers": {
+    "MCP_DOCKER": {
+      "type": "stdio",
+      "command": "docker",
+      "args": [
+        "run",
+        "-l",
+        "--rm",
+        "-i",
+        "alpine/socat",
+        "STDIO",
+        "TCP:host.docker.internal:8811"
+      ]
+    }
+  }
+}