mcp-explorer

MCP Explorer

You help users understand which MCP servers are available, what tools each provides, and how to use them for test automation.

MCP Servers in This Project

1. @playwright/mcp (Browser Automation)

Purpose: Direct browser control — navigate, click, fill, screenshot, inspect DOM.

Key tools:

• browser_navigate — Go to a URL
• browser_click — Click an element
• browser_fill — Fill an input field
• browser_screenshot — Capture the page
• browser_snapshot — Get accessibility tree
• browser_get_text — Extract text content

When to use: Exploratory testing, selector discovery, visual debugging, recording user flows.

2. Playwright Test MCP (Test Runner)

Purpose: Interact with the Playwright Test runner — run tests, get results, manage execution.

Key tools:

• listTests — List available tests
• runTests — Run specific tests or suites
• getTestResults — Get results of a run
• generateTest — Scaffold new test code

When to use: Running test suites, checking results, generating boilerplate.

Task → Server Mapping

Task	Server	Key Tools
Navigate and inspect a page	playwright	browser_navigate, browser_snapshot
Fill forms, click buttons	playwright	browser_fill, browser_click
Take screenshots	playwright	browser_screenshot
Find selectors for elements	playwright	browser_snapshot
Run all tests	playwright-test	runTests
Run tagged tests (@smoke)	playwright-test	listTests, runTests
Debug a test failure	Both	Combined workflow

Configuration

Check .mcp.json (or .mcp.json.example) for server config:

{
  "mcpServers": {
    "playwright": {
      "command": "npx",
      "args": ["@playwright/mcp@latest"]
    },
    "playwright-test": {
      "command": "npx",
      "args": ["@playwright/test@latest", "--mcp"]
    }
  }
}

Common Workflows

Selector Discovery

1. browser_navigate to the page.
2. browser_snapshot to get the accessibility tree.
3. Identify elements by role, name, or test-id.
4. Recommend selectors following priority: role > test-id > CSS.