home / mcp / on-demand scanning mcp server
MCP Server generated by mcp.ag2.ai
Configuration
View docs{
"mcpServers": {
"ag2-mcp-servers-on-demand-scanning-api": {
"command": "python",
"args": [
"mcp_server/main.py",
"stdio"
],
"env": {
"CONFIG": "{...} (JSON string configuration)",
"SECURITY": "API keys or other security parameters",
"CONFIG_PATH": "path/to/mcp_config.json"
}
}
}
}You run an MCP Server to expose a given API specification through a lightweight, protocol-powered endpoint. This server lets you connect your MCP client to the On-Demand Scanning API spec and interact with it in a stable, well-defined way, handling communication and protocol details for you.
To start and use the MCP Server, boot the local server process and connect your MCP client to it via the stdio transport. You can also configure security and runtime parameters through environment variables to tailor how the server authenticates and operates.
Steps to use the server after starting it: - Start the server in stdio mode using the provided command. - Point your MCP client at the running server, using the stdio channel as the transport. - Provide configuration through the supported environment variables or configuration file to control security and behavior. - Interact with the API via the MCP client, taking advantage of the server’s translation layer between the client and the OpenAPI-based specification.
Prerequisites you need before installing:
- Python 3.9 or newer
- pip and uv (optional for an alternative runner)Step by step installation flow: - Ensure Python 3.9+ is installed on your system. - Install the project in editable mode to include development dependencies (or install via uv if you prefer). - Install development dependencies to enable linting, static analysis, and tests.
# Use the editable install path to bring in development dependencies
pip install -e ".[dev]"
# Alternative with uv-based execution for editable mode
uv pip install --editable ".[dev]"Running the server relies on a single entry point that accepts stdio as the transport. You can start the server with a direct Python command and then use an MCP client to communicate.
Linting and formatting tool used to enforce code quality and style across the project.
Static type checker used to catch type errors before runtime.
Testing framework used to run unit and integration tests.
Pre-commit hooks configured to run linting, formatting, and basic checks before commits.
Build and publish tooling used to build and release the MCP server.