Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.sprig.com/llms.txt

Use this file to discover all available pages before exploring further.

Sprig MCP is currently in limited release, and product launch is expected in early June.For more information, reach out to Sprig at support@sprig.com or through your dedicated customer success team.

Connect Sprig via MCP

Use the Sprig MCP server to connect your survey data directly to AI tools- no manual exports needed. Your studies, responses, segments, and Agent Context become available as live context wherever you work.
Summary Gradient MCP
Before you start, ensure you have the following:
  1. A Sprig account with at least one study
  2. Admin access to enable MCP
  3. One of the supported AI clients below:
    • Claude.ai
    • Claude Code
    • Claude Cowork
    • ChatGPT
    • OpenAI Codex
    • Gemini CLI
    • GitHub CoPilot
    • Cursor
    • Linear
    • Notion

Step 1: Enable MCP in Sprig

  1. From the Integrations tab, navigate to AI & MCP
  2. Toggle Sprig MCP Server ON
  3. Copy your MCP server URL: https://mcp.sprig.com/mcp
Only administrators have permissions to turn Sprig MCP on

Step 2: Connect your AI client

Team and Enterprise plans. Owner role required.Admin setup
  1. From the sidebar in claude.ai, select Customize, then Connectors.
  2. Click the + icon, then Add Custom Connector.
  3. Give it the name “Sprig” and paste in the Sprig MCP server URL: https://mcp.sprig.com/mcp
  4. Click Connect and complete the OAuth authorization flow by logging in with your Sprig account.
Members can now find Sprig under Customize → Connectors and connect with one click.End user setupIf your admin has already added Sprig:
  1. From the sidebar in claude.ai, select Customize, then Connectors.
  2. Find “Sprig” in the list and click Connect.
  3. Complete the OAuth authorization flow by logging in with your Sprig account.
  4. Try: “Summarize the results from my last NPS study in Sprig.”
If you are setting up Sprig yourself (Free, Pro, Max plans):
  1. From the sidebar in claude.ai, select Customize, then Connectors.
  2. Click the + icon, then Add Custom Connector.
  3. Give it the name “Sprig” and paste in the Sprig MCP server URL: https://mcp.sprig.com/mcp
  4. Click Connect and complete the OAuth authorization flow by logging in with your Sprig account.
  5. Try: “Summarize the results from my last NPS study in Sprig.”
Tip: Claude can now read your Sprig studies, responses, and themes, no copy/paste needed.
Individual setup, Developers
  1. In your terminal, run claude mcp add sprig -t http https://mcp.sprig.com/mcp and follow the prompts.
  2. Run claude mcp list to verify the connection. Sprig should appear, but it likely will not be connected yet.
  3. Start a session to confirm the integration is live. The first time you invoke a Sprig tool, a browser window may open to complete the authorization flow. Try: “What is the name of my most recently created study?”
  4. Once connected, try asking: “What are the top themes from our most recent Sprig study?”
Tip: Pull user research context directly into your coding sessions. Ask what users said about the feature you’re building.
For Free, Pro, and Max users (or Team/Enterprise Owners adding for their org)Admin Setup
  1. In the Claude Cowork desktop app, go to Settings → Connected tools. From the sidebar, select “Customize,” then “Connectors.”
  2. Select Add custom connector. Give it a name (e.g., “Sprig”) and paste in the Sprig MCP server URL: https://mcp.sprig.com/mcp
  3. Click Connect to authorize the connection in a new browser tab. Once you’ve granted permissions, return to Claude.
Once added, individual members can enable it themselves via Customize > Connectors, where they’ll find and connect to the Sprig connector by name.For non-Owner Team/Enterprise membersEnd User Setup
  1. Go to Customize > Connectors and click the + button.
  2. Search for “Sprig” (or whatever your Owner named the connector) and select it.
  3. Click Connect to authorize your individual connection.
  4. Ask Cowork to pull Sprig study results into a doc, spreadsheet, or slide as part of any automated task.
Tip: Great for non-developers who want to automate research reporting. Cowork can pull Sprig data and write it directly into files.
Plus, Team, and Enterprise plans. Admin role required.End user setupIf your admin has already added Sprig:
  1. Go to chatgpt.comApps → Search for Sprig
  2. Click Connect and follow the prompts to authorize ChatGPT to access your Sprig account
  3. Sprig will now be available in the “+” menu under “More
Tip: Sprig survey data becomes available as context in any ChatGPT conversation or custom GPT you build.
Add Sprig to ChatGPT as described above under “ChatGPT”. Sprig will then be available as an app in Codex.
End users and developers
  1. Locate ~/.gemini/settings.json and add the Sprig server to the mcpServers list:
    {
      "mcpServers": {
        "sprig": {
          "httpUrl": "https://mcp.sprig.com/mcp",
          "oauth": {
            "enabled": true
          }
        }
      }
    }
    
  2. Restart the Gemini CLI and use the following command to authenticate with the Sprig MCP:
    /mcp auth sprig
    
  3. In the window that opens, allow all read access and click Authorize Gemini CLI MCP Client
  4. When successful, you will see an Authentication Successful screen
  5. Back in the Gemini CLI, you can verify that the Sprig MCP is connected properly by running the following command:
    /mcp list
    
Gemini CLI is ideal for scripting research workflows. Combine Sprig data with Google Workspace outputs like Docs and Sheets.
Admin SetupVS Code with the GitHub Copilot Chat extension is required (the older GitHub Copilot extension is deprecated, use GitHub.copilot-chat instead). VS Code version 1.99 or later is required.If your organization uses a Copilot Business or Enterprise plan, an org admin must also enable the “MCP servers in Copilot” policy in GitHub organization settings; it is disabled by default. Admins register the Sprig MCP server once. Developers in the org get it automatically and only need to complete the OAuth flow on first use, with no JSON editing required on their end.Option A: Shared workspace config (recommended)
Create a .vscode/mcp.json file in your project root and commit it to your repo:
{
  "servers": {
    "sprig": {
      "type": "http",
      "url": "https://mcp.sprig.com/mcp"
    }
  }
}
All developers who open the project in VS Code will have Sprig available automatically.Option B: User-level global configFor a personal setup that applies across all projects without modifying any repo, open the Command Palette (Cmd/Ctrl+Shift+P), run “MCP: Open User Configuration”, and add the following:
{
  "servers": {
    "sprig": {
      "type": "http",
      "url": "https://mcp.sprig.com/mcp"
    }
  }
}
Note: The key must be “servers”, not “mcpServers” or “github.copilot.chat.mcp.servers”. Using the wrong key will cause the server to silently fail to load.
Whether you use option A or Option B, each developer will see an OAuth prompt on first use and will need to login to Sprig to complete the connection.End User SetupIf your admin has deployed Sprig via shared workspace config or you’ve set it up via user-level config, the server is registered in your VS Code. You need to start it and authorize it on first use.
  1. Open your mcp.json file (either .vscode/mcp.json or your user config via the Command Palette → “MCP: Open User Configuration”). Look for the Start button that appears inline at the top of the "servers" block and click it. If you don’t see it, try opening Copilot Chat in Agent mode first, or run “MCP: List Servers” from the Command Palette to start it from there.
  2. Open Copilot Chat (Cmd/Ctrl+Alt+I).
  3. Switch to Agent mode using the mode dropdown in the chat input.
    Sprig tools are only accessible in Agent mode, and they will not appear in Ask or Edit mode.
  4. Click the tools icon in the top-left of the chat box to confirm Sprig appears in the tool list.
  5. Invoke a Sprig tool or ask a Sprig-related question. A browser window will open to complete the OAuth authorization flow.
  6. Log in with your Sprig account and grant access.
  7. Try: “What were the general themes from my latest Sprig study?”
    Tip: Useful for bringing user research into your code editor. Ask Copilot to reference Sprig feedback while reviewing or writing features.
Individual setup, Developers
  1. In Cursor, open a New Agent and run the following command to install mcp-remote:
    install mcp-remote
    
  2. Navigate to Settings → Tools & MCPs and click Add Custom MCP
  3. In the mcp.json file, add the following:
    {
      "mcpServers": {
        "sprig": {
          "command": "mcp-remote",
          "args": ["https://mcp.sprig.com/mcp"]
        }
      }
    }
    
  4. Open the menu in the top right and click Save File
  5. In the window that opens, login to Sprig and click Authorize Gemini CLI MPC Client
  6. When successful, you will see an Authentication Successful screen
  7. Open a New Agent and try: “List my Sprig studies”
Tip: Pull user research context while writing product code. Ask Cursor what users said about a specific feature.

Linear

Admin setupWorkspace admin role required.
  1. Go to Linear → Settings → Integrations → MCP.
  2. Paste in the Sprig MCP server URL: https://mcp.sprig.com/mcp
  3. Click Connect and complete the OAuth authorization flow by logging in with your Sprig account.
Members can now connect their individual accounts once the admin has enabled Sprig.End users
  1. Go to Linear → Settings → Integrations → MCP.
  2. Find Sprig and click Connect
  3. Complete the OAuth authorization flow by logging into Sprig
  4. Ask Linear’s AI assistant to pull feedback from Sprig when creating or triaging issues.
Tip: Close the loop between user feedback and product decisions. Create Linear issues directly from Sprig response themes.
Admin Setup
  1. Go to Notion → Settings → AI → Connected tools.
  2. Click Connect a tool and enter your Sprig MCP server URL and API key.
  3. Complete the OAuth authorization flow by logging into Sprig.
  4. Member can now connect to individual accounts once the admin has enabled Sprig.
  5. Use Notion AI to summarize Sprig study results directly into your research docs.
Tip: Ask Notion AI: “Summarize the key themes from our latest Sprig study and add them to this page.”

What Sprig exposes via MCP

DescriptionTool
Studies were formerly surveys.
This endpoint allows for the export of study configurations.
The name will eventually change to Study, but at this time survey will be
accepted anywhere study appears in the API.
This endpoint allows you retrieve up to 1000 surveys at a time.
https://api.sprig.com/v1/surveys
Use this endpoint to retrieve responses from your studies.
This endpoint allows you to retrievei up to 1000 responses at a time.
https://api.sprig.com/v1/responses
Use this endpoint to retrieve themes, along with the responses
that are associated with each theme, from your studies.
https://api.sprig.com/v1/themes

What you can do with Sprig MCP

Try these prompts in any connected AI client once Sprig MCP is set up.
  1. Summarize a study Summarize the key themes from my most recent Sprig study and highlight the top 3 issues users mentioned.
  2. Extract NPS insights Pull all NPS studies from Sprig in the last 30 days. What are the most common reasons users gave for low scores?
  3. Compare studies over time Compare the results from my Q1 and Q2 onboarding surveys in Sprig. What changed?
  4. Segment analysis From my latest Sprig study, show me how responses differ between power users and new users.
  5. Feature feedback digest Pull all Sprig responses that mention [feature name] and summarize what users are asking for.
  6. Draft a research summary doc Using my latest Sprig study results, write a one-page research summary I can share with my team.
  7. Turn feedback into hypotheses Based on the open-text responses in my latest Sprig study, generate 5 product hypotheses I could test next.
  8. Prioritize from feedback From my Sprig study on checkout experience, rank the top pain points by frequency and suggest which to fix first.
  9. Prep for a research readout Using my Sprig study results, create a bullet-point readout I can present in a 10-minute team meeting
  10. Look up by Study ID or Study Title
When both Sprig and another tool are connected to the same AI client, you can run workflows that span both, pulling data from Sprig and acting on it in another system automatically.**Sprig + Linear **Summarize the top 5 pain points from my latest Sprig study and create a Linear issue for each one, assigned to the product team.Sprig + Jira Pull negative responses from my Sprig checkout study and create Jira tickets with the user quotes as description text.Sprig + Notion Fetch my latest Sprig study results and write a research summary into my Notion research repository page.Sprig + Figma Pull the top usability issues from my Sprig study on the onboarding flow and add them as comments on the relevant Figma frames.Sprig + Slack Every time a Sprig study gets more than 50 responses, post a summary of the top themes to the #product-research Slack channel.Sprig + HubSpot Find all Sprig respondents who gave an NPS score below 6 and update their HubSpot contact record with a ‘detractor’ tag.Sprig + Salesforce Pull Sprig survey responses from enterprise-tier users and log the key feedback as activity notes on their Salesforce account records.Sprig + Google Sheets Export all responses from my latest Sprig study into a new Google Sheet, with one row per response and columns for question, answer, and segment.Sprig + Confluence Pull my last 3 Sprig studies and write a quarterly research digest page in Confluence summarizing key findings across all three.Sprig + GitHub From my Sprig developer experience study, extract bug reports and usability issues and create GitHub issues with appropriate labels.
Note: MCP to MCP workflows require both Sprig and the destination tool to be connected in the same AI client. Check that both connections are active before running these prompts.

Troubleshooting

  1. Connection not working Make sure you completed the OAuth authorization flow fully. Try disconnecting and reconnecting Sprig from your AI client’s connector settings, then log in with your Sprig account again when prompted.
  2. No studies appearing Make sure you have at least one published study in your workspace and that your account has access to it.
  3. OAuth window did not open Check that your browser is not blocking pop-ups. Try opening the authorization URL manually if your client provides one.
  4. Permission denied errors Your Sprig account may not have access to the studies you are querying. Check your workspace role with your Sprig admin.
  5. Config not recognized (Claude Code, Codex, Gemini CLI, GitHub Copilot) Make sure the config file is saved in the correct location and the JSON is valid. Restart the client after making changes.
  6. End users not seeing the Sprig connector (Claude, Claude Cowork, ChatGPT, Linear, Notion) The connector must be added by a workspace Owner or admin first. Ask your admin to follow the Admin setup steps for your tool above.
  7. GitHub Copilot: Sprig not appearing after admin deployment Make sure you are on VS Code 1.99 or later with the GitHub Copilot extension up to date. If using the workspace config option, confirm the .vscode/mcp.json file has been committed and pulled to your local repo.
Need help? Contact support@sprig.com or reach out to your dedicated customer success representatives.