microsoft/playwright-mcp

Powerful capabilities — review before trusting.

Environment variables (config / keys) 3

Shell / command execution 4

AI review

This is a legitimate Microsoft Playwright MCP server extension with no deceptive or malicious intent. The static findings show standard development patterns (CI checks, child_process usage in build scripts, environment variable access) that are expected for a browser automation tool. The release instructions and contributing guidelines are transparent about the project's purpose and workflow.

low No prompt injection or hidden instructions detected — All instructions in CLAUDE.md, release.md, and CONTRIBUTING.md are standard development workflow documentation for maintaining the Playwright MCP project. No attempts to manipulate the agent or exfiltrate data were found.
low Standard development patterns in static analysis — The flagged items (child_process usage, env access) are used in build scripts (roll.js, update-readme.js) and test files, which is expected for a Node.js project that needs to run shell commands for version rolling, README generation, and CI checks. These are not indicators of malicious behavior.
low No credential exfiltration or deceptive tool descriptions — The extension's stated purpose matches its actual functionality: providing browser automation via Playwright through MCP. No instructions to read or send secrets/credentials were found. Tool descriptions in the codebase are accurate and match the project's documented capabilities.

Model: deepseek-chat

Static findings

Environment variables (config / keys) · Reads environment variables (config / API keys)

low playwright.config.ts:24 forbidOnly: !!process.env.CI,

info tests/fixtures.ts:83 if (process.env.CI && process.platform === 'linux')

low update-readme.js:186 if (process.env.PRINT_ENV) {

Shell / command execution · Executes shell / system commands

medium roll.js:3 const { execSync } = require('child_process');

info tests/cli.spec.ts:16 import child_process from 'child_process';

info tests/library.spec.ts:16 import child_process from 'child_process';

medium update-readme.js:21 const { execSync } = require('child_process');

Scanning every extension your team installs?

Pro & Team add monitoring, private scans, and a CI gate for unsafe extensions.

MCPVet is a heuristic aid, not a security guarantee. A clean grade does not prove an extension is safe; always review code and instructions you don't trust.