MCP Hot Update Server Demo
About
A hot update server example based on the FastMCP framework that enables dynamic loading and updating of utility functions without restarting the server, managed via JSON configuration files.
Details
- Author
- whitejoce
- GitHub stars
- 1
- Downloads
- 161
- Categories
- Other
Jump to
- Dynamic loading of utility functions
- Manage tools via JSON configuration files
- Real-time hot update capability
- Detailed error handling and logging
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
MCP Hot Update Server DemoCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install Python 3.7+ and the fastmcp library with pip install fastmcp. Run the server with python server.py. Define tools in a tools.json file with name, description, and code fields.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"mcp hot update server demo": {
"hot-update-MCP-Server": {
"command": "python",
"args": [
"server.py"
]
}
}
}
}McpServers
{
"hot-update-MCP-Server": {
"command": "python",
"args": [
"server.py"
]
}
}MCP Hot Update Server Demo
<p align="center"> <a href="./README_CN.md">中文文档</a> | <a href="https://deepwiki.com/whitejoce/hot-update-MCP-Server">DeepWiki for this Repository</a> </p> This is a hot update server example based on the FastMCP framework, allowing dynamic loading and updating of utility functions without restarting the server.Features
- Supports dynamic loading of utility functions - Manage tools via JSON configuration files - Real-time hot update capability - Detailed error handling and loggingKey Code
``python
exec(tool_code, globals(), local_namespace)
if inspect.isfunction(obj) and obj.__name__ in local_namespace:
if found_func is not None:
# Handle potential duplicate function definition if needed
pass # Or log a warning, raise an error, etc.
found_func = obj
if found_func:
# Add the tool to MCP using the database name
mcp.tool(
name=tool_name, description=tool_description
)(found_func)
print(f"Successfully registered tool '{tool_name}'")
`
🚀 Quick Start
1️⃣ Environment Setup
Make sure you have Python 3.7+ installed. Then, install the required library:
`bash
$ pip install fastmcp
`
fastmcp: Python implementation library for MCP.
2️⃣ Run MCP Server
`bash
$ python server.py
`
The tools.json file format is as follows:
`json
[
{
"name": "Tool Name",
"description": "Tool Description",
"code": "def function_name(params):\n # Function implementation\n return result"
}
]
`
Security Notes
⚠️ Important Security Warning:
- The current implementation uses exec() to execute code read from the JSON file
- Additional security measures are required for production environments:
- Code signature verification
- Sandbox environment restrictions
- Code review mechanisms
- Access control
Development Suggestions
1. Test all utility functions in the development environment
2. Ensure the JSON file format is correct
3. Add appropriate error handling
4. Implement a code review process
5. Consider adding version control mechanisms
Workflow
`mermaid
graph TD
A[Start Server] --> B[Load tools.json]
B --> C{Parse JSON}
C -->|Success| D[Iterate Tool List]
C -->|Failure| E[Output Error Log]
D --> F[Execute Tool Code]
F --> G{Code Execution}
G -->|Success| H[Register to MCP]
G -->|Failure| I[Output Error Log]
H --> J[Server Running]
J --> K[Wait for New Requests]
``Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
