MCP Search Images

🚀 MCP Icon Search, Download, and Generation Guide

MCP offers a comprehensive toolset for quickly obtaining and generating icon resources during development. This document details how to use MCP for icon search, download, and custom generation.

🚀 Quick Start

1. Start the MCP service:

   python3 mcp_server.py
   

2. Integrate the MCP client into your development environment:

   from mcp_client import MPClient
   
   client = MPClient()
   

✨ Features

1. Icon Search

• Supported Sources:

  • Unsplash
  • Pixabay
  • Freepik
  • Noun Project

• Advanced Filtering:

  • Search by keywords
  • Filter by color
  • Filter by size
  • Filter by style (e.g., flat, skeuomorphic)

2. Icon Download

• Download Formats:

  • PNG
  • SVG
  • JPG

• Custom Naming:

  • Support specifying filenames
  • Support setting save paths

3. Icon Generation

• Supported Styles:

  • Flat design
  • Skeuomorphic design
  • Line drawing
  • Filled design

• Custom Attributes:

  • Color
  • Size
  • Outline
  • Background
  • Text content

💻 Usage Examples

Basic Usage

Icon Search

client.search_images(keywords="technology", source=Unsplash)

Icon Download

client.download_image(url="https://example.com/image.png", filename="tech-icon.png")

Icon Generation

client.generate_icon(
    description="Generate a blue technology icon",
    color="#00ff00",
    size=(256, 256),
    style=FLAT_STYLE
)

Advanced Usage

Batch Processing

• Batch Search:

  client.search_batch(keywords=["AI", "Machine Learning"], count_per_keyword=5)
  

• Batch Generation:

  for style in [FLAT_STYLE, OUTLINED_STYLE]:
      client.generate_icon(description="Email icon", style=style)
  

Custom Download Path

client.download_image(
    url="https://example.com/image.png",
    save_path="/Users/username/Desktop/images"
)

📚 Documentation

Error Troubleshooting

MCP Connection Issues

• Symptom: Encounter the "Fail to create client" error when adding the MCP service.
• Solutions:

  1. Check the service status:

     lsof -i :5173
     

     If there is no output, start the service:

     python3 mcp_server.py
     

  2. Test the connection:

     curl -v http://localhost:5173
     

  3. Modify the connection settings:

     • Ensure the correct connection type is selected: SSE
     • Try using the IP address instead of localhost: http://127.0.0.1:5173
     • Ensure the URL does not contain extra slashes, use http://localhost:5173 instead of http://localhost:5173/

  4. Configure using the ServerLink method:

     type: sse
     server_link: http://localhost:5173/sse
     

Image Processing Failures

• Symptom: Encounter exceptions when downloading or generating icons.
• Solutions:
  • Check the network connection.
  • Ensure the target server is running normally.
  • View detailed error logs.

Best Practices

Efficiency Optimization

1. Use Caching:

   client.enable_cache(max_size=1000)
   

2. Batch Process Tasks:

   from mcp_client import MPClient
   from concurrent.futures import ThreadPoolExecutor
   
   executor = ThreadPoolExecutor(max_workers=5)
   futures = []
   for keyword in keywords_list:
       future = executor.submit(client.search_images, keyword)
       futures.append(future)
   

Code Standards

1. Naming Conventions:

   • Use meaningful variable names.
   • Follow the project's coding style.

2. Exception Handling:

   try:
       client.generate_icon(description="Test icon")
   except Exception as e:
       print(f"Failed to generate icon: {e}")
   

3. Logging:

   import logging
   
   logging.basicConfig(
       level=logging.INFO,
       format='%(asctime)s - %(levelname)s - %(message)s'
   )
   

Advanced Features

Custom Plugins

• Develop a Plugin:

  def custom_search(self, keywords):
      # Custom search logic
      pass
  
  client.add_plugin(custom_search)
  

Batch Processing Tools

• Task Queue:

  python3 mcp_batch_processor.py --input_file=keywords.txt --output_dir=/path/to/output
  

Performance Tuning

Memory Optimization

1. Limit Cache Size:

   client.enable_cache(max_size=1000)
   

2. Process Data in Batches:

   from mcp_client import MPClient
   
   def process_batch(batch):
       for item in batch:
           # Processing logic
           pass
   
   client.process_in_batches(process_batch, batch_size=50)
   

Network Optimization

1. Use a Proxy:

   client.set_proxy("http://proxy.example.com:8080")
   

2. Limit Bandwidth:

   networkctl limit bandwith mcp-server 5mbit
   

Security Considerations

Authentication and Authorization

• Integrate OAuth:

  from oauthlib.oauth2 import OAuth2Session
  
  client = MPClient(
      auth=OAuth2Session(client_id="your_client_id", 
                        client_secret="your_client_secret")
  )
  

Data Encryption

1. Transport Layer Encryption:

   curl -k https://localhost:5173
   

2. Data Storage Encryption:

   from cryptography.fernet import Fernet
   
   key = Fernet.generate_key()
   cipher_suite = Fernet(key)
   encrypted_data = cipher_suite.encrypt(data.encode())
   

Appendix

Frequently Asked Questions

1. How to handle large amounts of data?

   • Use paginated search.
   • Enable the caching mechanism.
   • Configure a proxy server.

2. How to improve the quality of generated icons?

   • Provide high-resolution source files.
   • Use the SVG format.
   • Adjust generation parameters.

Reference Documents

1. MCP Official Documentation
2. Python Icon Processing Library

📄 License

This project is licensed under the MIT License. See the LICENSE file for specific terms.

Contribution Guide

1. How to contribute code?

   • Submit a Pull Request to the GitHub repository.
   • Follow the project's coding standards.
   • Participate in issue discussions.

2. How to report issues?

   • Submit an Issue to the GitHub repository.
   • Provide detailed error information.
   • Provide steps to reproduce the issue.