Verity
Start free
MCP Server

Generate & run tests from your IDE.

The Verity MCP server plugs into your AI coding agent — Cursor, VS Code, Claude Code, Windsurf — so you can generate, propose, and run Playwright tests against your localhost without leaving the editor. One-click install below.

01
Node.js 18+ (required)
The server runs via npx. Install from nodejs.org — a system install, then fully restart your IDE.
02
A compatible IDE
Cursor, VS Code, Claude Code, Windsurf, or any MCP-capable client.
03
Your project token
Free account → Dashboard → project → Local / MCP. It authenticates the server.

One click — VS Code will open and add the Verity MCP server. Then paste your project token when prompted.

Install in VS Code →

Or from the terminal:

code --add-mcp '{"name":"verity","command":"npx","args":["-y","@qashifthq/verity-mcp"],"env":{"VERITY_TOKEN":"YOUR_VERITY_TOKEN"}}'
Or add it manually (Command Palette)
  1. Open the Command Palette (⌘⇧P)
  2. Run MCP: Add Server
  3. Choose Command (stdio)
  4. Command to run: npx -y @qashifthq/verity-mcp
  5. Name it verity
  6. Pick the scope (workspace or global)
  7. Add env VERITY_TOKEN = your project token

Replace YOUR_VERITY_TOKEN with your project token from the dashboard (Local / MCP). Keep it secret — it's the same token your CI uses.

Getting spawn npx ENOENT?

That means your IDE can't find npx — GUI apps don't inherit your terminal's PATH (common if Node was installed via nvm). Fixes, in order:

  1. Install Node.js from nodejs.org (a system install, not only nvm), then fully quit and reopen your IDE.
  2. Still failing? Use the absolute path to npx. Find it with which npx in your terminal, then set that as the command:
{
  "mcpServers": {
    "verity": {
      "command": "/absolute/path/to/npx",
      "args": ["-y", "@qashifthq/verity-mcp"],
      "env": { "VERITY_TOKEN": "YOUR_VERITY_TOKEN" }
    }
  }
}

On macOS the path is usually /opt/homebrew/bin/npx (Apple silicon), /usr/local/bin/npx (Intel), or ~/.nvm/versions/node/vXX/bin/npx (nvm).

What it can do

  • Generate tests
    Describe a flow or point it at a URL — it proposes Playwright specs you review.
  • Run on localhost
    Execute tests against your local dev server, right from the agent.
  • Own the code
    Specs land in your repo — Playwright you keep, no lock-in.
  • Same token as CI
    Authenticates with your project token; results roll into your dashboard.

Need a token? Create a free accountand open your project's Local / MCP tab.