All servers / Browser / Playwright Stealth MCP server
Playwright Stealth MCP server
Local implementation of Playwright Stealth MCP server
Tested by ConnectATDScore 80 out of 100. Tested 2026-11-19 on package version 0.2.3. Package: playwright-stealth-mcp-server · source
Score breakdown
| Runs | 18 / 20 | |
|---|---|---|
| Tool descriptions | 20 / 20 | |
| Package safety | 20 / 20 | |
| Maintenance | 14 / 20 | |
| Adoption | 8 / 20 |
What we saw
- Starts and answers the MCP handshake: yes (4.5 seconds including download)
- Tools offered: 6
- Last published: 2026-06-14
- Weekly downloads (npm): 610
- License: MIT · Direct dependencies: 5
Want this server set up for your team?
ConnectATD builds AI automation and wires up MCP servers like this one. Tell us what you are trying to do.
Talk to ConnectATDTools and how they are described
Agents choose tools from these descriptions. Under each name you can see how many of its parameters are explained.
browser_execute2 of 2 parameters describedExecute Playwright code in the browser. This tool runs JavaScript code with access to a Playwright `page` object. The browser session persists across calls, so you can navigate, interact with elements, and extract data across multiple tool invocations. **Available in scope:** - `page` - The Playwright Page object with full API access **Example usage:** Navigate to a page: ```javascript await page.goto('https://example.com'); return await page.title(); ``` Click an element and wait: ```javascript await page.click('button.submit'); await page.waitForSelector('.success-message'); ``` Extract text content: ```javascript const items = await page.$$eval('.item', els => els.map(el => el.textContent)); return items; ``` Fill a form: ```javascript await page.fill('input[name="email"]', 'test@example.com'); await page.fill('input[name="password"]', 'secret'); await page.click('button[type="submit"]'); ``` **Returns:** - `success`: boolean indicating if execution succeeded - `result`: JSON stringified return value (if any) - `error`: error message (if failed) - `consoleOutput`: array of console messages from the page **Note:** When STEALTH_MODE=true, the browser includes anti-detection measures to help bypass bot protection.
browser_screenshot4 of 4 parameters describedTake a screenshot of the current page, a specific element, or a page region. Captures the visible viewport, full page, a specific element, or a rectangular region as a PNG image. Screenshots are saved to filesystem storage and can be accessed later via MCP resources. **Parameters:** - `fullPage`: Whether to capture the full scrollable page (default: false) - `selector`: CSS selector of a specific element to screenshot (e.g., '#main-content', '.hero-banner', 'table.results') - `clip`: Region of the page to screenshot as {x, y, width, height} in pixels - `resultHandling`: How to handle the result: - `saveAndReturn` (default): Saves to storage AND returns inline base64 image - `saveOnly`: Saves to storage and returns only the resource URI (more efficient for large screenshots) **Note:** `fullPage`, `selector`, and `clip` are mutually exclusive. Only one can be specified per call. **Returns:** - With `saveAndReturn`: Inline base64 PNG image data plus a resource_link to the saved file - With `saveOnly`: A resource_link with the `file://` URI to the saved screenshot **Dimension Limits:** Full-page screenshots are limited to 8000 pixels in any dimension. If a full-page screenshot would exceed this limit, it is automatically clipped from the top-left corner and a warning is included in the response. Element and clip screenshots are not subject to this limit. **Use cases:** - Verify page state after navigation - Screenshot a specific element like a chart, table, or form - Capture a region of the page by coordinates - Debug automation issues - Store screenshots for later reference via MCP resources
browser_get_stateno parametersGet the current browser state. Returns information about the current browser session including the URL, page title, and whether a browser is open. **Returns:** - `currentUrl`: Current page URL - `title`: Current page title - `isOpen`: Whether a browser session is active - `stealthMode`: Whether stealth mode is enabled - `headless`: Whether running in headless mode - `proxyEnabled`: Whether a proxy is configured - `ignoreHttpsErrors`: Whether HTTPS certificate errors are being ignored
browser_closeno parametersClose the browser session. Shuts down the browser and clears all state. A new browser will be launched on the next execute call.
browser_start_recordingno parametersStart recording the browser session as a video. This tool begins capturing all browser interactions as a WebM video file. It works by recycling the browser context with video recording enabled. **Session state preservation:** Cookies and localStorage are automatically saved and restored when the context is recycled. sessionStorage for the current origin is also preserved on a best-effort basis. In rare cases involving multiple origins, some state may not be restored — if you experience authentication issues, log in again. **Behavior when already recording:** If recording is already active, the current recording will be stopped (saving the video) and a new recording will begin. **The video is NOT available until you call `browser_stop_recording`.** Playwright only finalizes the video file when the recording context is closed. **Returns:** - Confirmation that recording has started - The URL the browser was previously on (if any) — the browser navigates back to it automatically
browser_stop_recordingno parametersStop recording the browser session and save the video. This tool stops the active video recording, saves the video file, and returns a resource URI for the recorded video. **Session state preservation:** Cookies and localStorage are automatically saved and restored when the context is recycled. sessionStorage for the current origin is also preserved on a best-effort basis. The browser automatically navigates back to the URL it was on before the recording stopped. **Error:** Returns an error if no recording is currently active. Use `browser_start_recording` first. **Returns:** - A resource_link with the `file://` URI to the saved video file (WebM format) - The video can be accessed later via MCP resources
Tested in a sandbox with an empty environment: no API keys, a throwaway folder, nothing from a real account. A passing test shows the server runs and responds. It is not a security audit.