Frequently Asked Questions¶

General Questions¶

What is RapidTriageME?¶

RapidTriageME is an AI-powered browser debugging platform that captures browser data (console logs, network requests, errors, screenshots) and makes it available to AI assistants like Claude and ChatGPT through the Model Context Protocol (MCP). This enables intelligent debugging assistance and automated problem analysis.

How does RapidTriageME work?¶

The system consists of three main components:

Chrome Extension - Captures browser data using DevTools API
Browser Connector - Local server that aggregates and serves data
MCP Server - Bridges browser data to AI assistants

Data flows from web pages → extension → connector → MCP → AI assistant, enabling real-time debugging analysis.

Is RapidTriageME free to use?¶

Yes, RapidTriageME is open source software released under the MIT License. You can use, modify, and distribute it freely. The core functionality is completely free.

What browsers are supported?¶

Currently supported: - ✅ Chrome (primary support) - ✅ Chromium-based browsers (Edge, Brave, etc.) - ⚠️ Firefox (experimental support) - ❌ Safari (not supported due to extension API limitations)

What AI assistants work with RapidTriageME?¶

Any AI assistant that supports the Model Context Protocol (MCP): - ✅ Claude Desktop (native support) - ✅ Claude API (via MCP client) - ✅ Custom MCP clients - 🔄 ChatGPT (via MCP integration - coming soon)

Installation & Setup¶

How do I install RapidTriageME?¶

Choose your preferred installation method:

Quick Install (Recommended)From SourceDocker

# Install browser connector
npm install -g @yarlisai/rapidtriage-server

# Install MCP server  
npm install -g @yarlisai/rapidtriage-mcp

# Start the system
rapidtriage-server

# Clone repository
git clone https://github.com/yarlisai/rapidtriage.git
cd rapidtriage

# Run setup script
./run.sh all

# Pull and run container
docker run -p 3025:3025 yarlisai/rapidtriage:latest

How do I load the Chrome extension?¶

Download the extension from releases or build from source
Open Chrome and go to chrome://extensions/
Enable "Developer mode" (toggle in top right)
Click "Load unpacked" and select the extension directory
Verify the extension appears in your extensions list

How do I configure Claude Desktop?¶

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "rapidtriage": {
      "command": "rapidtriage-mcp",
      "args": ["--config", "/path/to/mcp-config.json"]
    }
  }
}

Config file locations: - Mac: ~/Library/Application Support/Claude/claude_desktop_config.json - Windows: %APPDATA%\Claude\claude_desktop_config.json - Linux: ~/.config/Claude/claude_desktop_config.json

Usage & Features¶

What types of browser data can be captured?¶

RapidTriageME captures comprehensive browser debugging information:

Console Logs: console.log(), console.warn(), console.error(), etc.
JavaScript Errors: Uncaught exceptions, syntax errors, runtime errors
Network Requests: XHR, Fetch API, resource loading (with headers & responses)
Screenshots: Full page or viewport screenshots on demand
DOM Elements: Selected element details, attributes, styling
Performance Metrics: Load times, resource timing, Lighthouse audits
Page Information: URL, title, metadata, viewport dimensions

How do I capture a screenshot?¶

Multiple ways to capture screenshots:

Through AI Assistant: "Take a screenshot of the current page"
DevTools Panel: Click "Capture Screenshot" button
API Call: POST http://localhost:3025/capture-screenshot
Command Line: curl -X POST http://localhost:3025/capture-screenshot

Can I access browser data remotely?¶

Yes! RapidTriageME supports remote access through Cloudflare Workers:

Deploy the Cloudflare Worker component
Configure custom domain (e.g., rapidtriage.me)
Generate JWT authentication tokens
Access browser data from anywhere securely

How do I clear captured data?¶

Several methods to clear data:

Through AI: "Clear all logs and data"
DevTools Panel: Click "Wipe Logs" button
API Call: POST http://localhost:3025/wipelogs
Restart Server: Data clears on server restart

What's the difference between local and remote access?¶

Feature	Local Access	Remote Access
Performance	Fastest (localhost)	Slower (network latency)
Security	Most secure	Requires authentication
Accessibility	Same machine only	Available anywhere
Setup	Simple	Requires Cloudflare setup
Cost	Free	May incur Cloudflare costs

Troubleshooting¶

The extension shows "Failed to connect"¶

Common solutions:

Check server status:
```
curl http://localhost:3025/.identity
```
Verify server is running:
```
ps aux | grep rapidtriage-server
```
Check port availability:
```
lsof -i :3025
```

Restart server:

pkill -f rapidtriage-server
rapidtriage-server

Console logs aren't appearing¶

Check these common issues:

DevTools must be open - Extension only captures when DevTools is active
Refresh the page - After loading extension, refresh target page

Check debugger attachment:

// In extension console
chrome.debugger.getTargets(console.log);

Screenshots fail to capture¶

Verify these conditions:

Valid tab - Must be http:// or https:// page (not chrome://)
Active tab - Tab must be currently active/visible
Permissions - Extension needs activeTab permission
Downloads folder - Must be writable for file saving

MCP server not connecting to Claude¶

Check config file path - Must be absolute path
Validate JSON syntax:
```
cat claude_desktop_config.json | jq .
```
Restart Claude Desktop - Close completely and reopen
Check server logs:
```
rapidtriage-mcp --log-level debug
```

High memory usage¶

Implement data limits to reduce memory consumption:

Limit log entries:

// In extension settings
maxLogEntries: 1000,  // Default: 50
maxNetworkEntries: 500  // Default: 50

Enable auto-cleanup:

// Clear data older than 5 minutes
autoCleanupInterval: 300000

Restart extension periodically - Reload extension to clear memory

Performance & Optimization¶

How can I improve performance?¶

Reduce data retention limits:

{
  "maxLogEntries": 100,
  "maxNetworkEntries": 50,
  "stringLengthLimit": 1000
}

Enable request filtering:

// Filter out noise (images, fonts, etc.)
filterResourceTypes: ['image', 'font', 'stylesheet']

Use caching:

# Enable response caching
rapidtriage-server --enable-cache --cache-ttl=30

What are the system requirements?¶

Minimum Requirements: - OS: Windows 10, macOS 10.15, Ubuntu 18.04+ - RAM: 2GB available memory - Browser: Chrome 88+, Edge 88+ - Node.js: v16+ (for local installation)

Recommended: - RAM: 4GB+ available memory - Storage: 1GB free space for logs/screenshots - Network: Stable internet for remote features

How much bandwidth does it use?¶

Bandwidth usage depends on data volume:

Local usage: Minimal (localhost only)
Remote access: 1-10MB per debugging session
Screenshots: 100KB-2MB per image
Heavy debugging: Up to 50MB/hour with full capture

Security & Privacy¶

Is my browser data secure?¶

Yes, RapidTriageME prioritizes security:

Local by default - Data stays on your machine
No external tracking - No analytics or telemetry
Encrypted remote access - HTTPS/WSS for remote connections
JWT authentication - Secure token-based access
Data sanitization - Automatic removal of sensitive patterns

What data is collected?¶

RapidTriageME only collects debugging data you explicitly capture:

Console output from web pages you debug
Network requests from those pages
JavaScript errors occurring on those pages
Screenshots when you request them
Page metadata (URL, title, etc.)

Not collected: - Personal browsing history - Data from other tabs/windows - System information beyond browser context - Credentials or sensitive form data

Can I use RapidTriageME in production?¶

Local development: ✅ Yes, safe and recommended Production monitoring: ⚠️ Use with caution - Only capture non-sensitive applications - Avoid pages with user data/credentials
- Consider privacy implications - Use data filtering and sanitization

How do I redact sensitive data?¶

Configure automatic data sanitization:

{
  "dataSanitization": {
    "enabled": true,
    "redactPatterns": [
      "password",
      "token",
      "api_key",
      "secret"
    ],
    "redactRegex": [
      "\\b\\d{4}-\\d{4}-\\d{4}-\\d{4}\\b",  // Credit cards
      "\\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\\.[A-Z|a-z]{2,}\\b"  // Emails
    ]
  }
}

Development & Customization¶

Can I customize RapidTriageME?¶

Yes! RapidTriageME is highly customizable:

Configuration files - Modify behavior through JSON config
Custom filters - Add filtering rules for data capture
API extensions - Add custom endpoints to browser connector
MCP tools - Create custom tools for AI integration
Source code - Full source available for modification

How do I add custom data capture?¶

Extend the Chrome extension:

// In content script
function captureCustomData() {
  const customData = {
    type: 'custom-event',
    data: {
      // Your custom data
    },
    timestamp: Date.now()
  };

  // Send to background script
  chrome.runtime.sendMessage(customData);
}

Can I integrate with other AI services?¶

Yes, through the MCP protocol or custom integrations:

MCP Integration: Any MCP-compatible AI service
Custom API: Build adapters for specific AI APIs
Webhook Integration: Send data to external services
SDK Development: Create libraries for your preferred language

How do I contribute to development?¶

We welcome contributions! Here's how to get started:

Fork the repository on GitHub

Set up development environment:

git clone your-fork.git
cd rapidtriage
npm install
npm run dev

Make your changes and test thoroughly
Submit a pull request with clear description
Follow coding standards and include tests

See our Contributing Guide for detailed guidelines.

Advanced Usage¶

How do I set up load balancing?¶

For high-traffic scenarios, use multiple browser connector instances:

# Start multiple instances on different ports
rapidtriage-server --port 3025 &
rapidtriage-server --port 1422 &
rapidtriage-server --port 1423 &

# Configure load balancer (nginx example)
upstream rapidtriage {
    server localhost:3025;
    server localhost:1422;
    server localhost:1423;
}

Can I use RapidTriageME in Docker?¶

Yes! Docker support is available:

# Pull official image
docker pull yarlisai/rapidtriage:latest

# Run with port mapping
docker run -p 3025:3025 -p 1422:1422 yarlisai/rapidtriage:latest

# Or build from source
docker build -t rapidtriage .
docker run -p 3025:3025 rapidtriage

How do I monitor multiple browser instances?¶

Configure unique session IDs for each browser:

// Extension configuration
const sessionConfig = {
  sessionId: `browser-${Math.random().toString(36).substr(2, 9)}`,
  browserInstance: window.navigator.userAgent
};

Can I export debugging data?¶

Yes, multiple export options available:

# Export as JSON
curl http://localhost:3025/export/json > debug-session.json

# Export as CSV
curl http://localhost:3025/export/csv > debug-session.csv

# Export logs only
curl http://localhost:3025/console-logs > console-logs.json

Deployment & Production¶

How do I deploy to production?¶

Follow our deployment guides:

Local Production - Self-hosted setup
Cloudflare Deployment - Global edge deployment
Docker Installation - Containerized deployment

What's the recommended architecture?¶

Small Team (1-5 developers):

Browser Extension → Local Connector → MCP → Claude Desktop

Medium Team (5-20 developers):

Multiple Extensions → Shared Connector → MCP Server → Team AI Access

Large Team/Enterprise:

Extensions → Load Balanced Connectors → Cloudflare Worker → Global AI Access

How do I monitor system health?¶

Use built-in health monitoring:

# Health check endpoints
curl http://localhost:3025/health
curl http://localhost:1422/health

# System metrics
curl http://localhost:3025/metrics

# Performance monitoring
curl http://localhost:3025/stats

Set up alerts for production use:

# Example monitoring script
#!/bin/bash
HEALTH=$(curl -s http://localhost:3025/health | jq -r '.status')
if [ "$HEALTH" != "healthy" ]; then
    echo "RapidTriage health check failed!" | mail -s "Alert" [email protected]
fi

Getting Help¶

Where can I get support?¶

Multiple support channels available:

📖 Documentation: rapidtriage.me/docs
🐛 GitHub Issues: Bug reports and feature requests
💬 Discord: Real-time community support
📧 Email: [email protected] for enterprise support
🌟 Stack Overflow: Tag questions with rapidtriage

How do I report a bug?¶

When reporting bugs, include:

System information (OS, browser, versions)
Steps to reproduce the issue
Expected vs actual behavior
Error messages and console logs
Screenshots if relevant
Configuration files (redact sensitive data)

Use our bug report template on GitHub for best results.

Is commercial support available?¶

Yes! Enterprise support options:

🏢 Enterprise License: Priority support and SLA
🛠 Custom Development: Feature development and integration
📋 Training: Team training and best practices
☁️ Managed Hosting: Fully managed cloud deployment

Contact [email protected] for details.

How can I stay updated?¶

Stay informed about updates:

⭐ Star the GitHub repo for release notifications
📧 Join mailing list for major announcements
🐦 Follow @YarlisAI on Twitter for updates
📱 Join Discord for real-time updates and discussions

Still have questions?¶

If you didn't find your question here:

Search the docs - Use the search function above
Check GitHub Issues - Your question might already be answered
Ask the community - Join our Discord or Stack Overflow
Contact support - For urgent issues or enterprise questions

Quick Links: - Getting Started Guide - Installation Instructions - Troubleshooting Guide - API Documentation