For the complete documentation index, see llms.txt. This page is also available as Markdown.

Set up the Dscout MCP

The Dscout MCP is currently in closed beta. Only certain users have access, and the feature may change before general availability. For more information, contact your Dscout Account Director.

The Dscout MCP lets you use Dscout research methods and analysis tools in AI workflows. Your data stays secure.

Supported tools

You can connect the Dscout MCP to any AI tool that allows custom connectors/MCPs. If you're unsure whether an AI tool you use allows this, work with your organization's internal IT team to verify.

Security measures

The Dscout MCP includes these security measures:

  • OAuth sign-in ensures no API keys or tokens are shared.

  • A Dscout account keeps a team member in the loop.

  • Customizable read/write permissions let you choose what the Dscout MCP can do.

For supported tasks, see Dscout MCP tools.

Dscout's MCP endpoint

Point your MCP client at the following endpoint to leverage the Dscout MCP:

https://app.dscout.com/astro/mcp

Connect your AI tool

For tool-specific guidance on configuring the Dscout MCP, see the sections below:

To set up the Dscout MCP in Claude:

1

Create a custom connector named Dscout.

2

Enter the Dscout MCP URL: https://app.dscout.com/astro/mcp

3

Click Add.

Now, the Dscout MCP is available as a custom connector in Claude. You'll be prompted to authenticate using your standard Dscout credentials.

To set up the Dscout MCP in Claude Code:

1

Run the following command in your CLI:

2

Run claude in your CLI to open Claude.

3

Run /mcp. A list of your enabled MCPs is displayed, with Dscout showing as Needs authentication.

4

Select Dscout and press Enter.

The authentication page opens. Grant the MCP access to your Dscout account, loggin in with your standard Dscout credentials if prompted.

To set up the Dscout MCP in Figma Make:

1

Create a custom connector named Dscout.

2

Enter https://app.dscout.com/astro/mcp as the MCP server URL.

3

Click Connect.

Now, the Dscout MCP is available as a custom connector in Figma Make. You'll be prompted to authenticate using your standard Dscout credentials.

To set up the Dscout MCP in Codex:

1

Add a custom MCP server named Dscout.

2

Enter https://app.dscout.com/astro/mcp as the URL.

3

Click Save. You're returned to your list of Codex MCPs.

4

Click Authenticate beside Dscout.

The authentication page opens. Grant the MCP access to your Dscout account, logging in with your standard Dscout credentials if prompted.

To set up the Dscout MCP in Cursor:

1

Add a custom MCP server named Dscout.

2

Enter https://app.dscout.com/astro/mcp as the Server URL.

3

Click Add MCP.

4

Locate Dscout in your list of custom Cursor MCPs.

5

Click Login.

The authentication page opens. Grant the MCP access to your Dscout account, logging in with your standard Dscout credentials if prompted.

To set up the Dscout MCP in Antigravity:

1

Follow the Antigravity instructions for connecting custom MCP servers and add the following to your mcp_config.json file:

2

Save the configuration and restart Antigravity.

Note: You will see an error in Antigravity at this point in the configuration. This is expected and will be resolved by following the steps below.

3

Navigate to the Antigravity Settings menu. Then, select Customizations.

4

Beside Dscout, click Authenticate. The Dscout authentication page is displayed.

5

Click Approve to grant Antigravity access to your Dscout account. An authentication code is displayed.

6

Copy and paste the authentication code into Antigravity. Then, click Submit to complete authentication.

7

Click Refresh.

Now, the Dscout MCP is connected and ready for use in Antigravity.

To set up the Dscout MCP in VS Code:

1

Create a .vscode/mcp.json file in your workspace:

2

Open the Command Palette (Cmd+Shift+P / Ctrl+Shift+P) and run MCP: List Servers .

3

Start the Dscout server and complete the OAuth flow when prompted.

To configure the Dscout MCP across all workspaces, run MCP: Open User Configuration from the Command Palette and add the server configuration there.

To set up the Dscout MCP in Microsoft Copilot:

1

Go to the Tools page for your agent.

2

Select Add a tool.

3

Select New tool.

4

Select Model Context Protocol. The MCP onboarding wizard is displayed.

5

Fill in the following fields.

  • Server name: Dscout.

  • Server description: Give the agent read and write access to the Dscout research platform, including studies, screeners, participants, sessions, transcripts, highlights, and responses. Use it to create and configure studies, manage participants, pull transcripts and responses, and search past research for findings and highlights.

  • Server URL: https://app.dscout.com/astro/mcp

6

Select OAuth 2.0 as the authentication method. Then, select Dynamic discovery.

7

Click Create.

Now, the MCP is available for your Copilot agents, and you'll be prompted to authenticate using your Dscout credentials upon first use.

To set up the Dscout MCP in Windsurf (Devin):

1

Add the following to Windsurf's mcp_config.json file:

2

Save the file, completing authentication when prompted.

Note: You may be prompted to grant Windsurf permission to open a window in your browser. You must grant this permission as this is how you will connect the tool to your Dscout account.

Now, the Dscout MCP is available in Windsurf.

Any MCP-compatible client can connect over streamable HTTP.

Point your client to the following endpoint:

No authentication fields are required during configuration. Users will be prompted to authenticate with their standard Dscout credentials upon their first use.

Note: Users must have an active Dscout account to use the Dscout MCP. They will be prompted to authenticate in the browser using their existing Dscout credentials on their initial use of the MCP.

Verify the connection

After connecting to the Dscout MCP, verify the connection with one of these prompts:

What studies are currently launched across my Dscout projects? You'll see a list of launched studies, including parent project, study type, number of entries, and launch date.

How many responses have I received in my most recently launched study? You'll see the number of submitted responses for the study you launched most recently.

Show me the most recent entry submitted to my [study]. You'll see the most recent response for your requested study, including the participant's name, submission date, and key takeaways from their response data.

What was one common thread across responses in my [study]? You'll see an underlying theme that appeared across participant responses in your requested study, including supporting quotes and participant names.

Create a new Dscout usability test for me. You'll be prompted to enter some basic details about your study (study name, desired project, etc.) and to answer questions about your research objectives.

If your AI tool returns an error, see Troubleshoot the Dscout MCP for help resolving it.

Learn more

Learn more about the Dscout MCP in these articles:

Last updated