StackMCP

QA / Testing Stack for Claude Code

Intermediate5 servers34.3K tokens

Configuration

{
  "mcpServers": {
    "playwright-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@playwright/mcp"
      ]
    },
    "puppeteer-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-puppeteer"
      ]
    },
    "sentry-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "sentry-mcp"
      ],
      "env": {
        "SENTRY_AUTH_TOKEN": "YOUR_SENTRY_AUTH_TOKEN"
      }
    },
    "github-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-github"
      ],
      "env": {
        "GITHUB_PERSONAL_ACCESS_TOKEN": "YOUR_GITHUB_PERSONAL_ACCESS_TOKEN"
      }
    },
    "filesystem-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-filesystem",
        "/path/to/allowed/directory"
      ]
    }
  }
}

CLI Commands

Alternatively, add each server via the Claude Code CLI:

claude mcp add playwright-mcp -- npx -y @playwright/mcp
claude mcp add puppeteer-mcp -- npx -y @modelcontextprotocol/server-puppeteer
claude mcp add sentry-mcp -e SENTRY_AUTH_TOKEN=YOUR_SENTRY_AUTH_TOKEN -- npx -y sentry-mcp
claude mcp add github-mcp -e GITHUB_PERSONAL_ACCESS_TOKEN=YOUR_GITHUB_PERSONAL_ACCESS_TOKEN -- npx -y @modelcontextprotocol/server-github
claude mcp add filesystem-mcp -- npx -y @modelcontextprotocol/server-filesystem /path/to/allowed/directory

Where to save

Paste the config above into:

~/.claude.json

Environment Variables

Replace the YOUR_ placeholders with your actual values.

SENTRY_AUTH_TOKENrequired

Sentry auth token

Used by: Sentry MCP

GITHUB_PERSONAL_ACCESS_TOKENrequired

GitHub personal access token

Used by: GitHub MCP

What’s in this stack

Playwright MCP20 tools

Automate browser interactions, take screenshots, fill forms, and test web applications using Microsoft Playwright from your AI editor.

The gold standard for browser automation. Write, run, and debug end-to-end tests with full page interaction from your AI assistant.

Puppeteer MCP8 tools

Automate browsers with Puppeteer. Navigate pages, take screenshots, fill forms, and generate PDFs from your AI editor.

A second browser automation engine for cross-tool coverage. Some legacy test suites depend on Puppeteer — having both avoids rewrites.

Sentry MCP8 tools

Query Sentry for error tracking, performance monitoring, and issue management directly from your AI editor.

Pull real production errors into your test planning. Write regression tests for the exact bugs your users are hitting.

GitHub MCP20 tools

Access the GitHub API to manage repositories, issues, pull requests, branches, and workflows directly from your AI editor.

Link test results to PRs, manage CI workflows, and ensure tests pass before merging. QA lives in the PR review cycle.

Filesystem MCP11 tools

Read, write, search, and manage files on your local filesystem with secure directory-scoped access for your AI editor.

Read test fixtures, write test reports, and manage configuration files. QA workflows are file-heavy by nature.

Other editors

CursorVS CodeWindsurfClaude DesktopContinue