@visualq/mcp

MCP server for VisualQ — run VRT, poll results, and read structured failure reports from Cursor, Claude Desktop, or any MCP client.

Quick start (Cursor)

Add to .cursor/mcp.json or global MCP settings:

{
  "mcpServers": {
    "visualq": {
      "command": "npx",
      "args": ["-y", "@visualq/mcp"],
      "env": {
        "VISUALQ_API_KEY": "vq_live_…",
        "VISUALQ_BASE_URL": "https://visualq.ai"
      }
    }
  }
}

Create an API key in VisualQ: Project → Settings → API keys (prefix vq_live_).

Tools

Phase 1: list_projects, list_scenarios, run_vrt, run_baseline, get_run_status, wait_for_run, get_run_failures, get_diff_stats, get_quality_score

Phase 2: get_run_history, get_scenario_details, compare_runs, check_setup_health, explain_vrt_failure, perf_get_latest_report, seo_get_report, a11y_get_report, tracking_get_plan

Development

npm install
npm run build
VISUALQ_API_KEY=vq_live_… node dist/index.js

Sync tool manifest from the visualq backend repo:

cd ../visualq && npm run mcp:export-manifest

Repository

Published from abecms/visualq-mcp. Backend routes live in abecms/visualq (/api/mcp/v1/invoke).