MCP Integration

Connect UserTold.ai to Claude, Cursor, or any MCP-compatible agent for direct tool calls, resource reads, and prompts.

Connect

Add to your MCP client config:

Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "usertold": { "type": "http", "url": "https://mcp.usertold.ai/mcp" }
  }
}

Claude Code (.mcp.json in project root or ~/.claude/mcp.json):

{
  "mcpServers": {
    "usertold": { "type": "http", "url": "https://mcp.usertold.ai/mcp" }
  }
}

After adding the config, the client opens a browser login (OAuth 2.1 + PKCE). No manual token handling.

Once connected, call initialize then tools/list to see the full tool catalog.

projectRef

Most tools are scoped to a project. Accepted formats:

  • orgHandle/projectHandle — e.g. acme/billing
  • Project ID — e.g. proj_abc123

What you can do

Tools are grouped by domain:

  • projects — create and manage projects, check signal health and readiness
  • studies — design and activate interview studies from scripts
  • sessions — list sessions, read transcripts, trigger reprocessing
  • signals — list, annotate, dismiss, and link extracted signals to tasks
  • tasks — create tasks from signal clusters, review evidence, push to GitHub/Linear, measure impact
  • screeners — manage screeners and qualification rules

Research loop

A typical agent workflow:

  1. Check project signal health to understand coverage gaps
  2. Design and activate a study
  3. Sessions run automatically as users interact with the widget
  4. List and curate extracted signals (annotate, dismiss, link)
  5. Create evidence-backed tasks from signal clusters
  6. Review evidence, merge duplicates, and push tasks to GitHub or Linear
  7. Re-run sessions after shipping — compare signal rates to measure impact

See also