Automate Trading Operations with MCP for Trade Server

Trade Server runs a brokerage's daily work: accounts, orders, routing, margin, and reporting. What no interface can ship with is every question a desk will ask. A report built around how one desk defines exposure. Two sets of figures that no single screen shows together. The same check applied across a set of accounts. Requirements like these often lead to a development request, a manual export, or another specialist tool.

Your Bourse has released MCP for Trade Server to help brokers automate trading operations through natural-language interaction with the Trade Server backend. Brokers and prop firms can connect a compatible AI assistant of their choice to Your Bourse Trade Server and use natural-language instructions to query authorised data and initiate permitted actions. It is available now for brokers running Trade Server.

The connection is not limited to reading data. A risk manager can ask for total exposure on EURUSD or for every account in negative equity without requesting a separate report. Subject to the connected user's permissions, a dealer can initiate a hedge or close a position by instruction, while operations teams can complete authorised account and group actions in the same way. The assistant reads current Trade Server data at the time of the request.

How brokers can use MCP for Trade Server

1. Connect an AI assistant. A broker or prop firm connects a compatible AI assistant or agent of its choice to Trade Server using existing credentials.
2. Apply the existing access scope. The permissions attached to those credentials determine which data and actions the connected user can access. MCP does not expand them.
3. Ask a question or issue an instruction. The user communicates in natural language instead of navigating through multiple screens, exporting data, or requesting a new report.
4. Review the response or proposed action. Trade Server returns current authorised data. Where an instruction affects positions or funds, the assistant first returns a preview without executing it.
5. Confirm the action. Once a person approves it, the action goes ahead and is recorded in Trade Server just as if it had been done through the interface.

Practical brokerage use cases

1. Run ad hoc client reports. A dealer or risk manager can ask for the most profitable clients for the day without building a separate report or extracting the data through the Manager API.
2. Investigate trading queries. A client complaint such as a profitable position closing at a loss can be investigated against the relevant Trade Server journal and trading data, reducing the need to manually gather information across the system.
3. Cross-reference account data. Teams can request accounts above a chosen balance threshold and compare them against their groups or books in the same workflow.
4. Create tailored risk views. Brokers can use Trade Server data through MCP to build risk views and dashboards around their own operational requirements rather than relying only on predefined reports.

The connection uses the brokerage's existing Trade Server credentials and does not require a separate custom AI integration. Brokers can use their preferred LLM, like Claude or ChatGPT to connect, and MCP for Trade Server can run on the brokerage's own infrastructure.

Trade Server clients can also extend MCP-enabled access to their own trading clients, giving them a simpler route to connect API-driven tools, develop AI-powered trading strategies, and interact with the broker's trading infrastructure. The broker defines the access and permissions available to each client.

“MCP for Trade Server makes the Trade Server API easier to use in day-to-day operations. Instead of developing and testing a script for every request, teams can describe what they need in natural language and work directly with the Trade Server data and functions available to them,” said Valter Timanov, Chief Product Officer, Your Bourse.

Watch Valter Timanov, Chief Product Officer at Your Bourse, walk through MCP for Trade Server on YouTube and demonstrate how natural-language instructions can be used across real Trade Server workflows:

Book a Demo to see how MCP could work with your Trade Server setup: https://bit.ly/4qfMlpp