userken.
Documentation

Connect userken. to Claude

userken. runs as an MCP server. Connect once and the tools show up alongside everything else in Claude.

01 · Claude Desktop

Add to your config.

Edit ~/Library/Application Support/Claude/claude_desktop_config.json on macOS or the equivalent on Windows.

{
  "mcpServers": {
    "userken": {
      "url": "https://userken.com/mcp"
    }
  }
}

Restart Claude Desktop. The userken. tools appear in the tools panel.

02 · Claude.ai web

Add the integration.

  1. Project → Settings → Integrations
  2. Add MCP server
  3. URL: https://userken.com/mcp
  4. Save. The tools become available in any chat in this project.
03 · Tools

What Claude can call.

ToolDescription
search_reviewsFilter reviews by query, rating, theme, platform, or publication.
semantic_search_reviewsVector search via local embeddings — find reviews by concept, not keyword.
find_similar_reviewsKeyword fallback when embeddings aren't available.
get_persona_quotesReal quotes that represent how a persona talks.
get_all_personasAll personas across publications.
get_personaFull details for a specific persona.
get_reviews_for_personaReviews assigned to a given persona cluster.
get_theme_summaryStats for a theme: count, sentiment, rating distribution.
get_statsOverall review stats.
trigger_syncManually pull the latest reviews.
get_sync_statusLast sync time and result.
04 · Focus group prompts

Pre-built panels you can drop in.

Each prompt is a complete system message for Claude — paste it as a project instruction or as the first turn of a conversation. The personas use userken.'s tools to back every response with real review quotes.

05 · Example asks

Quick prompts to try.

Concept search

"Find reviews where users complain about getting blocked from articles after a free trial across all publications."

Focus group

"Run a focus group with the Reuters personas: would they pay $2 more for an ad-free tier?"

Competitive comparison

"Compare what users complain about most in Bloomberg vs WSJ — break down by theme and sentiment."

Theme stats

"How has notification sentiment changed for the BBC News app over the past six months?"