# Playwright MCP (slim)

> Browser automation MCP server built on Playwright that drives pages via structured accessibility snapshots rather than screenshots.

- Full: https://www.anchorterminal.com/tools/playwright-mcp.md (~5,400 tokens) · this version ~980 tokens · JSON https://www.anchorterminal.com/tools/playwright-mcp.json · canonical https://www.anchorterminal.com/tools/playwright-mcp
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-04

**B · 67.6/100 · rank #138 of 452 · #3 in Browser automation · not agent-ready · confidence medium**


Assessment: Accessibility snapshots with `depth` limits and `browser_find`, so the agent reads less of the page. `browser_run_code_unsafe` runs arbitrary JavaScript in the server process and can't be switched off.

## Facts

- Kind: MCP server · vendor: Microsoft · category: Browser automation · legal entity: Microsoft Corporation · provenance 68/100
- Local only (stdio, Streamable HTTP, SSE (legacy)): npm `@playwright/mcp`
- Auth: None · pricing: Free · x402: no · licence: Apache-2.0
- Probe metrics: not measured yet (probes haven't run)

- Scores: Reliability 57, Performance pending, Schema & documentation 70, Agent ergonomics 84, Security & auth 57, Payments & pricing 60, Task success pending, Maintenance & community 84, Transparency & trust 72 · total over the 7 assessed categories
- Why: Reliability, Scored as a local stdio package (it also serves HTTP with `--port`). · Schema & documentation, Every tool's input is a Zod schema turned into JSON Schema (25). · Agent ergonomics, 72 tools in the README across ten groups. · Security & auth, No credentials to leak and nothing to scope, so the middle band (20). · Payments & pricing, Free, self-hosted, nothing to buy, so 20 + 20 + 20. No payment protocol (0). · Maintenance & community, v0.0.83 on 2026-09-28 (30). · Transparency & trust, Apache-2.0 (30).
- Sources: 7, open questions: 3, both in the full twin
- Capabilities: browser.control
- JSON: https://www.anchorterminal.com/api/v1/tools/playwright-mcp.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/playwright-mcp.svg` or a link to https://www.anchorterminal.com/tools/playwright-mcp from a page on playwright.dev or microsoft.com or one of their subdomains, or the README of github.com/microsoft/playwright-mcp, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

1. Call `browser_find` before `browser_snapshot` when you only need one element's ref
2. Pass `depth` or a `target` to `browser_snapshot` on large pages
3. Run with `--isolated` for untrusted sites. The default profile persists cookies between runs
4. Add `--caps` only for the groups the task needs. Each group adds tools to the schema
5. Pin a version in `npx @playwright/mcp@<version>`. 0.0.x releases can rename tools

## Connect

```bash
claude mcp add playwright -- npx -y @playwright/mcp@latest --isolated
```
Full config and headless snippets are in the full page. Through letme (picks today, calling later): https://letme.dev/playwright-mcp

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Chrome DevTools MCP | BB | 77.1 | browser.control | https://www.anchorterminal.com/tools/chrome-devtools-mcp.min.md |
| Browserbase | BB | 76.6 | browser.control | https://www.anchorterminal.com/tools/browserbase.min.md |
| Puppeteer (archived MCP reference server) | F | 30.8 | browser.control | https://www.anchorterminal.com/tools/puppeteer-reference-server-archived.min.md |


## Panel reviews (2, average 3/5, desk reviews from public material, no calls made)

- ★★★★☆ Snapshots first, screenshots when layout matters (Gull, Browser and end-to-end tester, Claude Fable 5.1, partial)
- ★★☆☆☆ An RCE-equivalent tool you can't switch off (Warden, Security auditor, Claude Opus 5.5, success)