--- title: "Devin Desktop" description: "Devin Desktop is an AI coding assistant that can connect various Large Language Models (LLM) to the QuantConnect MCP Server." canonical: https://www.quantconnect.com/docs/v2/ai-assistance/mcp-server/devin-desktop markdown: https://www.quantconnect.com/docs/v2/ai-assistance/mcp-server/devin-desktop.md language: python source: https://github.com/QuantConnect/Documentation/tree/master/09%20AI%20Assistance/08%20MCP%20Server/30%20Devin%20Desktop modified: 2026-10-28 index: https://www.quantconnect.com/llms.txt --- # Devin Desktop [Documentation](https://www.quantconnect.com/docs/v2.md) > [AI Assistance](https://www.quantconnect.com/docs/v2/ai-assistance.md) > [MCP Server](https://www.quantconnect.com/docs/v2/ai-assistance/mcp-server.md) > Devin Desktop > Code examples are shown in Python. Append `?lang=csharp` to this URL for the C# version of this page. ## Introduction Devin Local is an AI coding assistant in Devin Desktop (formerly Windsurf) that can connect various Large Language Models (LLM) to the [QuantConnect MCP Server](https://www.quantconnect.com/docs/v2/ai-assistance/mcp-server/key-concepts.md). This page explains how to set up and use the server in Devin Desktop. > Devin Local truncates large responses from MCP servers, so it can be unstable for long backtests. ## Getting Started To connect the Devin Local agent in Devin Desktop to the QC MCP Server, follow these steps: 1. Install and open [Devin Desktop](https://devin.ai/download). 2. In the top navigation bar of Devin Desktop, click **View > Extensions**. 3. In the **Search Exensions in Marketplace** field, enter **QuantConnect**. 4. Click **Install**. 5. In the left navigation bar, click the ![QuantConnect icon in Cursor](https://cdn.quantconnect.com/i/tu/cursor-qc-icon.png) **QuantConnect** icon. 6. In the Initialization Checklist panel, click **Login to QuantConnect**. ![Initialization checklist in Devin Desktop, where you can log in to QuantConnect](https://cdn.quantconnect.com/i/tu/windsurf-initialization-checklist.png) 7. In the Devin Desktop window, click **Open**. ![Pop-up that enables you to open a QC link to login to your account](https://cdn.quantconnect.com/i/tu/windsurf-login-to-quantconnect.png) 8. On the Code Extension Login page, click **Grant Access**. 9. In the Select Workspace panel, click **Pull Organization Workspace**. ![Select workspace](https://cdn.quantconnect.com/i/tu/select-workspace.png) 10. In the Pull QuantConnect Organization Workspace window, click the cloud workspace ([organization](https://www.quantconnect.com/docs/v2/cloud-platform/organizations.md)) that you want to pull. ![Pull cloud organization](https://cdn.quantconnect.com/i/tu/pull-cloud-workspaces-1.png) 11. In the Pull QuantConnect Organization Workspace window, create a directory to serve as the organization workspace and then click **Select**. ![Pull cloud workspace](https://cdn.quantconnect.com/i/tu/pull-cloud-workspaces-2.png) It takes a few minutes to create a new organization workspace directory and populate it with the [the initial file structure](https://www.quantconnect.com/docs/v2/local-platform/development-environment/organization-workspaces.md#07-Directory-Structure). After the organization workspace is populated with the initial file structure, it pulls your cloud project files. 12. Press **Ctrl+Shift+P** to open the Command Palette, enter **Devin: Devin MCP Registry**, and then press **Enter**. 13. On the Devin Customizations page, click **Add custom MCP**. 14. In your `~ / AppData / Roaming / devin / mcp_config.json` file, add the following configuration: ``` { "mcpServers": { "qc-mcp": { "disabled": false, "url": "https://www.quantconnect.com/api/v2/mcp" } } } ``` 15. Save and close the `~ / AppData / Roaming / devin / mcp_config.json` file. 16. On the Devin Customizations page, click the qc-mcp server. ![The list of installed MCP servers in Devin Desktop](https://cdn.quantconnect.com/i/tu/devin-open-qc-mcp.png) 17. On the qc-mcp page, click **Connect**. ![The qc-mcp with an arrow pointing to the Connect button](https://cdn.quantconnect.com/i/tu/devin-connect-qc-mcp.png) 18. In the Authorize MCP Server form, select the organization you want to connect to and then click **Authorize**. ![QuantConnect form to Authorize MCP Server connection](https://cdn.quantconnect.com/i/tu/authorize-mcp-server.png) Devin Desktop should now show all the tools available through the QuantConnect MCP Server. ![List of QC MCP Server tools in Devin Desktop](https://cdn.quantconnect.com/i/tu/devin-qc-mcp-server-tools.png) 19. In Devin Desktop, [create a new project](https://www.quantconnect.com/docs/v2/local-platform/projects/getting-started.md#03-Create-Projects) or [open an existing one](https://www.quantconnect.com/docs/v2/local-platform/projects/getting-started.md#04-Open-Projects). 20. Press **Ctrl+L** to open Devin Local and begin interacting with the agent. To backtest and live trade with the local [Deployment Target](https://www.quantconnect.com/docs/v2/local-platform/key-concepts/deployment-targets.md), install Docker Desktop. The installation process depends on if your machine uses [Windows](https://www.quantconnect.com/docs/v2/local-platform/installation/install-on-windows.md#03-Install-Docker), [macOS](https://www.quantconnect.com/docs/v2/local-platform/installation/install-on-macos.md#03-Install-Docker), or [Linux](https://www.quantconnect.com/docs/v2/local-platform/installation/install-on-linux.md#03-Install-Docker). ## Models Devin Desktop supports several LLMs that you can use in the **Devin Local** panel, including GPT, Claude, and Gemini. To change the model, click the model name at the bottom of the **Devin Local** panel and then click the name of the model to use. ![Arrow clicking the model name in the Devin Local panel of Devin Desktop, showing the list of available models.](https://cdn.quantconnect.com/i/tu/windsurf-models.png) To view all the available models, see [AI Models](https://docs.devin.ai/desktop/models) in the Devin documentation. ## Quotas There are no quotas on the QuantConnect API, but Devin Desktop and the LLMs have some. To view the quotas, see [Quota-Based Usage](https://docs.devin.ai/desktop/accounts/quota) in the Devin documentation and see the quotas of the model you use. ## Troubleshooting The following sections explain some issues you may encounter and how to resolve them. ### Service Outages The MCP server relies on the QuantConnect API and the client application. To check the status of the QuantConnect API, see our [Status](https://status.quantconnect.com/) page. To check the status of Devin Desktop, see the [Devin Status](https://status.devin.ai/) page. To check the status of the LLM, see its status page. For example, Claude users can see the [Anthropic Status](https://status.anthropic.com/) page. ## Examples The following examples demonstrate the MCP server with Devin Local in Devin Desktop. ### Example 1: Hello World To test the server and client are working and connected, enter the following prompt into the Devin Local chat: > > Read the open project. The agent should call the `read_open_project` tool.