BrowserLoopSAFE
A Model Context Protocol (MCP) server for taking screenshots and reading console logs of web pages using Playwright.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://github.com/mattiasw/browserloop/actions/workflows/ci.yml) [](https://www.npmjs.com/package/browserloop) [](https://www.npmjs.com/package/browserloop)
⚠️ ARCHIVED: This project is archived and will not receive any further updates. With the release of Chrome DevTools MCP, a dedicated MCP server for browser automation is no longer necessary as that project provides more comprehensive browser interaction capabilities including screenshots, console monitoring, and much more.
A Model Context Protocol (MCP) server for taking screenshots and reading console logs from web pages using Playwright. This tool allows AI agents to automatically capture screenshots and monitor browser console output for debugging, testing, and development tasks.
NOTE: Almost all of the code in this repository has been auto-generated. That means you should probably not trust it too much. That being said, it does work and I'm using it myself.
NOTE: If the documentation is incorrect, please let me know or send a PR. If you too want to use a code generation tool to update the code for this project, PROJECT_CONTEXT.md has been used as context to give a good overview of the various parts of the project. It might be a bit messy now but it's a good starting point and you're welcome to update it.
Features
- 📸 High-quality screenshot capture using Playwright
- 📝 Console log monitoring and collection from web pages
- 🌐 Support for localhost and remote URLs
- 🍪 Cookie-based authentication for protected pages
- 🐳 Docker containerization for consistent environments
- ⚡ PNG, JPEG, and WebP format support with configurable quality
- 🛡️ Secure non-root container execution
- 🤖 Full MCP protocol integra
077e27312da7OBSERVED · 2026-10-09Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add browserloop -- npx -y [email protected]
{
"mcpServers": {
"browserloop": {
"command": "npx",
"args": [
"-y",
"[email protected]"
]
}
}
}Exposed tools (2)
2 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
read_console | read | |
screenshot | read |
Trust audit
SAFEgrade B · trust 87/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (2 observation(s))
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- found
Findings (8)
| **API Keys** | `sk_live_1234567890abcdef...` | `[API_KEY_MASKED]` |
name: 'eval(',join(currentDir, '../../package.json'), // From dist/src/
import { ScreenshotService } from '../../src/screenshot-service.js';} from '../../src/test-utils.js';
import type { Cookie } from '../../src/types.js';import { ScreenshotService } from '../../src/screenshot-service.js';@modelcontextprotocol/sdk, @types/sharp, playwright, sharp, zod, @biomejs/biome, @types/assert, @types/node
Gates applied: no_behavioural_pass.
077e27312da7full audit observations/trust-audit/mcp-server/mattiasw__browserloop.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-09 | 077e27312da7 | SAFE | B | 87 | first audit |
Questions
What is the BrowserLoop MCP server?
A Model Context Protocol (MCP) server for taking screenshots and reading console logs of web pages using Playwright.
What tools does BrowserLoop expose?
2 in total: 2 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is BrowserLoop safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (87/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does BrowserLoop need?
No credential environment variables were found in its source, so it appears to need none.
How does BrowserLoop run?
It speaks stdio, so it runs as a local process your client starts. It is published on npm as browserloop at 1.1.1.
How current is this page?
The grade is for one exact copy of the source (077e27312da7), read on 2026-10-09. The repository is watched and re-audited when it changes.