BrowserLoop
A Model Context Protocol (MCP) server for taking screenshots of web pages using Playwright, allowing AI agents to automatically capture and analyze screenshots for UI verification or other tasks.
README
BrowserLoop
A Model Context Protocol (MCP) server for taking screenshots of web pages using Playwright. This tool allows AI agents to automatically capture and analyze screenshots for UI verification or other 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
- 🌐 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 integration with AI development tools
- 🔧 Configurable viewport sizes and capture options
- 📱 Full page and element-specific screenshot capture
- ⚡ TypeScript with Biome for fast development
- 🧪 Comprehensive testing with Node.js built-in test runner
Quick Start
📦 NPX Usage (Recommended)
The easiest way to get started - no installation required!
# Install Chromium browser (one-time setup)
npx playwright install chromium
# Test that BrowserLoop works
npx browserloop@latest --version
That's it! The latest version of BrowserLoop will be downloaded and executed automatically. Perfect for MCP users who want zero-maintenance screenshots.
MCP Configuration
Add BrowserLoop to your MCP configuration file (e.g. ~/.cursor/mcp.json):
{
"mcpServers": {
"browserloop": {
"command": "npx",
"args": ["-y", "browserloop@latest"],
"description": "Screenshot capture server for web pages using Playwright"
}
}
}
💡 Using @latest ensures you always get the newest features and bug fixes automatically.
🚀 One-Click Install for Cursor
Add BrowserLoop to Cursor with a single click using this deeplink:
This deeplink will automatically configure BrowserLoop in your Cursor MCP settings with the optimal configuration using npx and the latest version.
Prerequisites: Make sure you have Chromium installed first:
npx playwright install chromium
Browser Installation Requirements
🚨 Critical: BrowserLoop requires Chromium to be installed via Playwright before it can take screenshots.
First-Time Setup (All Users)
Install Chromium browser:
npx playwright install chromium
Verify installation:
# Check Playwright installation
npx playwright --version
# Test BrowserLoop (if using NPX)
npx browserloop@latest --version
🐳 Docker Alternative
For containerized environments:
# Pull and run with Docker
docker run --rm --network host browserloop
# Or use docker-compose for development
git clone <repository-url>
cd browserloop
docker-compose -f docker/docker-compose.yml up
💻 Development Installation
For contributors or advanced users who want to build from source:
# Clone the repository
git clone <repository-url>
cd browserloop
# Install dependencies
npm install
# Install Playwright browsers (required for screenshots)
npx playwright install chromium
# OR use the convenient script:
npm run install-browsers
# Build the project
npm run build
MCP Configuration for Development
{
"mcpServers": {
"browserloop": {
"command": "node",
"args": [
"/absolute/path/to/browserloop/dist/src/index.js"
],
"description": "Screenshot capture server for web pages using Playwright"
}
}
}
Replace /absolute/path/to/browserloop/ with your actual project path.
Basic Usage
Once configured, you can use natural language commands in your AI tool:
Take a screenshot of https://example.com
Take a screenshot of https://example.com with width 1920 and height 1080
Take a screenshot of https://example.com in JPEG format with 95% quality
Take a full page screenshot of https://example.com
Take a screenshot of http://localhost:3000 to verify the UI changes
🔐 Cookie Authentication
BrowserLoop supports cookie-based authentication for capturing screenshots of login-protected pages during development:
Take a screenshot of http://localhost:3000/admin/dashboard using these cookies: [{"name":"connect.sid","value":"s:session-id.signature","domain":"localhost"}]
📖 For cookie extraction methods and development workflows, see:
Common development use cases:
- Local development servers with authentication
- Staging environment testing
- API documentation tools (Swagger, GraphQL Playground)
- Custom web applications during development
- Admin panels and protected routes
Documentation
- 🔐 Cookie Authentication Guide - Complete guide for authenticated screenshots
- 📚 Complete API Reference - Detailed parameter documentation, examples, and response formats
Key API Parameters
| Parameter | Type | Description | Default |
|---|---|---|---|
url |
string | Target URL to capture (required) | - |
width |
number | Viewport width (200-4000) | 1280 |
height |
number | Viewport height (200-4000) | 720 |
format |
string | Image format (webp, png, jpeg) | webp |
quality |
number | Image quality (1-100) | 80 |
fullPage |
boolean | Capture full page | false |
selector |
string | CSS selector for element capture | - |
📖 See docs/API.md for complete parameter details, usage examples, and configuration options.
Configuration
BrowserLoop can be configured using environment variables:
Basic Configuration
| Variable | Default | Description |
|---|---|---|
BROWSERLOOP_DEFAULT_WIDTH |
1280 |
Default viewport width (200-4000) |
BROWSERLOOP_DEFAULT_HEIGHT |
720 |
Default viewport height (200-4000) |
BROWSERLOOP_DEFAULT_FORMAT |
webp |
Default image format (webp, png, jpeg) |
BROWSERLOOP_DEFAULT_QUALITY |
80 |
Default image quality (0-100) |
BROWSERLOOP_DEFAULT_TIMEOUT |
30000 |
Default timeout in milliseconds |
BROWSERLOOP_USER_AGENT |
- | Custom user agent string |
Authentication Configuration
| Variable | Default | Description |
|---|---|---|
BROWSERLOOP_DEFAULT_COOKIES |
- | Default cookies as file path or JSON string (see Cookie Authentication Guide) |
Performance & Reliability
| Variable | Default | Description |
|---|---|---|
BROWSERLOOP_RETRY_COUNT |
3 |
Number of retry attempts for failed operations |
BROWSERLOOP_RETRY_DELAY |
1000 |
Delay between retries in milliseconds |
Logging & Debugging
| Variable | Default | Description |
|---|---|---|
BROWSERLOOP_DEBUG |
false |
Enable debug logging to /tmp/browserloop.log |
BROWSERLOOP_ENABLE_METRICS |
true |
Enable error metrics collection |
BROWSERLOOP_DISABLE_FILE_WATCHING |
false |
Disable automatic cookie file monitoring |
Debug Logging
When BROWSERLOOP_DEBUG=true, detailed logs are written to /tmp/browserloop.log including:
- Cookie file loading and automatic refresh events
- File watching status and recreation events
- Screenshot operation details
- Configuration changes and errors
Monitor logs in real-time:
tail -f /tmp/browserloop.log
Note: Logs are written to a file (not console) to maintain compatibility with MCP's stdio protocol.
Example MCP Configuration with Default Cookies
Method 1: JSON File (Recommended)
Create a cookies file:
// ~/.config/browserloop/cookies.json
[
{
"name": "connect.sid",
"value": "s:your-dev-session.signature",
"domain": "localhost"
}
]
Reference in MCP config:
{
"mcpServers": {
"browserloop": {
"command": "node",
"args": ["dist/src/mcp-server.js"],
"env": {
"BROWSERLOOP_DEFAULT_COOKIES": "/home/username/.config/browserloop/cookies.json",
"BROWSERLOOP_DEFAULT_FORMAT": "webp",
"BROWSERLOOP_DEFAULT_QUALITY": "85"
}
}
}
}
Method 2: JSON String (Legacy)
{
"mcpServers": {
"browserloop": {
"command": "node",
"args": ["dist/src/mcp-server.js"],
"env": {
"BROWSERLOOP_DEFAULT_COOKIES": "[{\"name\":\"session_id\",\"value\":\"your_session_value\",\"domain\":\"example.com\"},{\"name\":\"auth_token\",\"value\":\"your_auth_token\"}]",
"BROWSERLOOP_DEFAULT_FORMAT": "webp",
"BROWSERLOOP_DEFAULT_QUALITY": "85"
}
}
}
}
Troubleshooting
Common Issues
"Executable doesn't exist" Error
# Install Chromium browser (most common fix)
npx playwright install chromium
MCP Server Not Starting
- Test manually:
npx browserloop@latest --version - Verify requirements:
- Node.js 20+:
node --version - npm:
npm --version - npx:
npx --version
- Node.js 20+:
- Check MCP config JSON syntax
Screenshots Show Login Pages
- Use cookie authentication (see Cookie Authentication Guide)
- Check cookie expiration and domain settings
Network/Connection Issues
- Test with external URLs first:
https://example.com - For localhost: ensure your dev server is running
- Check firewall settings
Updating BrowserLoop
- NPX: Automatically uses latest version with
@latest- no manual updates needed! - Check current version:
npx browserloop@latest --version
Quick Diagnosis
# Test complete setup
node --version && npm --version
npx playwright --version
# Test BrowserLoop
npx browserloop@latest --version
Enable debug logging:
Set BROWSERLOOP_DEBUG=true in your MCP config and monitor /tmp/browserloop.log
📖 See docs/API.md#error-handling for detailed troubleshooting.
License
BrowserLoop is licensed under the GNU Affero General Public License v3.0 or later (AGPL-3.0-or-later).
What this means:
- ✅ Free to use - Personal and commercial use allowed
- ✅ Free to modify - You can adapt the code to your needs
- ✅ Free to distribute - Share copies with others
- ✅ Patent protection - Contributors provide patent grants
- ⚠️ Copyleft - Derivative works must also be open source under AGPL-3.0
- ⚠️ Network clause - If you run a modified version on a server, you must provide source code to users
For Network Services
Important: If you modify BrowserLoop and run it as a network service (e.g., web app, API server, or cloud service), the AGPL requires you to:
- Offer the complete source code to all users of your service
- Include a prominent notice about how users can access the source
- Use a compatible license for the entire service
License Files
- LICENSE - Full license text
Commercial Use
Organizations can use BrowserLoop under the AGPL for commercial purposes, but must comply with the copyleft requirements. If you need to keep modifications private, consider:
- Using BrowserLoop without modifications
- Contributing improvements back to the community
- Contacting the maintainers about potential alternative licensing arrangements
For questions about licensing, please open an issue or contact the maintainers.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.