Skip to main content

​
1. Obtain the Connection

Open QuantX, select Volatility Scaling Lab, and open Connection. Configure its URL and authorization using MCP Client Setup. Keep the access key in your client’s secret storage or environment, not in shared configuration.

​
2. Discover the Tools

Connect with an MCP client and list tools. Optionally call load_skills with {"request": {}} for the server’s workflow guidance. It is guidance, not an authorization step.

​
3. Calculate Volatility

Call compute-ewma-volatility-series with supplied decimal returns. This example needs no market-data provider:
Here 0.01 means a 1% return. The tiny sample checks integration only; it is not a useful risk estimate for trading.

​
4. Read the Result

Inspect the MCP tool’s structured content. Small payloads arrive inline under result. A large one is replaced by a stored reference, marked with data_stored: true:
Only then call result_get on the same server, passing that id:
The full value is in result.payload of that response. Check for tool errors before extracting it. A result_id that sits next to inline data is just a label for the call — it does not mean you have to fetch anything.

​
5. Export

Use Shared MCP Tools to save an artifact and read its bytes.