- PANW AIRS MCP Server
PANW AIRS MCP Server
PANW AI Runtime Security MCP Server
The PANW AI Runtime Security MCP Server allows you to integrate Palo Alto Networks AI security analytics capabilities directly into any client that supports the Model Context Protocol (MCP), such as Claude Desktop, Visual Studio Code, or Cursor IDE.
With this server, you can assess content security risks in real time when using large language models, ensuring security compliance for AI interactions.
🚀 Features
- Real-time content risk analysis
- Seamless integration with any MCP-compatible client
- Supports multiple input types (text input, output, code input, code output)
- Content detection based on Palo Alto Networks AI Security API
- Provides local version and SSE server version to adapt to different usage scenarios
🔑 Preparation
Before you start, you need:
- Palo Alto Networks AI Security API key
- Configure AI Security profile name
📋 Installation guide
Download project files
https://github.com/zm1990s/panw/tree/main/panw-mcp-server
⚙️ Usage
Use the local version in Cursor IDE
- Open Settings → MCP tab
- Click "Add new global MCP server"
- Paste the following configuration in the
mcp.jsonfile:
{
"mcpServers": {
"PANW-AI-Security": {
"command": "python",
"args": [
"~/ai/cursor/panw-mcp-server/local-version/src/panw_airs_mcp_local.py"
],
"env": {
"AIRS_API_KEY": "APIKEY",
"AIRS_PROFILENAME": "PROFILENAME"
}
}
}
}
Start SSE version locally
There are currently two versions of python files. If v2 is used, the key will be obtained directly from the http request without setting the environment variable. For details, see Call 3 in Dify.
- Set environment variables
export AIRS_API_KEY="your API key"
export AIRS_PROFILENAME="your profile name"
- Run the server:
python sse-version/src/panw_airs_mcp_sse.py --host 0.0.0.0 --port 8080
- The server will start on http://localhost:8080
- Configure your MCP client to connect to this URL
For more documentation, see: https://github.com/zm1990s/panw/blob/main/panw-mcp-server/README-EN.md
Server Config
{
"mcpServers": {
"PANW-AI-Security": {
"command": "uv",
"args": [
"run",
"src/panw_airs_mcp_local.py"
],
"env": {
"AIRS_PROFILENAME": "PROFILENAME",
"AIRS_API_KEY": "APIKEY"
}
}
}
}