Vexstorm Web Relay for Cloud AI Clients

The Vexstorm Web Relay is a secure HTTPS tunnel that connects cloud AI clients like ChatGPT to your local GameCatalyst installation.

What is the Vexstorm Web Relay?

It’s a managed relay service hosted by Vexstorm Studios. Your local GameCatalyst connects outbound to the relay server, which then provides a stable HTTPS URL that cloud AI clients can access. Think of it like a reverse SSH tunnel, but specifically designed for MCP.

When to use Vexstorm Web Relay:

• You’re using ChatGPT, Claude.AI Online, or Devin
• You want a stable URL that persists across sessions
• You don’t want to manage your own tunnel infrastructure
• You need outbound-only connections (no inbound firewall rules)

Benefits of Vexstorm Web Relay:

• Stable URLs — Your relay URL stays the same across sessions (e.g., https://webrelay.vexstorm.com/relay/abc123/mcp)
• No firewall configuration — Outbound-only connection, no inbound ports needed
• Managed infrastructure — Vexstorm Studios handles server maintenance and uptime
• One-click start/stop — Control from Dashboard Status tab or WhatIs site

How it works:

1. Click “Start” on the Web Relay card in the Dashboard Status tab
2. GameCatalyst connects outbound to wss://webrelay.vexstorm.com/ws/tunnel
3. The relay server assigns you a stable tunnel ID based on your project
4. Your relay URL becomes available (shown in Dashboard and WhatIs)
5. Cloud AI clients connect to this URL
6. Messages tunnel through the relay to your local GameCatalyst

Privacy and security:

• The relay does NOT inspect, store, or log your MCP messages
• It’s a pure tunnel — messages pass through encrypted
• Rate limiting prevents abuse (burst: 100ms min, window: 120/60s, budget: 10K daily)
• Your project data never touches the relay server

Required for ChatGPT:

ChatGPT requires a public HTTPS URL to connect MCP servers. The Vexstorm Web Relay provides this without exposing your local machine to inbound connections.

Stable URLs explained:

Unlike Cloudflare Quick Tunnel (which generates a new random URL each time), the Vexstorm Web Relay gives you a deterministic URL based on your project. This means your relay URL stays the same across sessions. You configure your AI client once and it keeps working every time you start the relay. No need to update your ChatGPT connector URL after every restart.

Rate limiting for safety:

The relay includes rate limiting to prevent abuse: a burst limit (minimum time between requests), a window limit (maximum requests per time period), and a budget limit (daily and monthly caps). These limits are generous enough for normal use but prevent runaway requests from causing problems. If you hit a rate limit, the relay returns a clear message explaining what happened and when you can retry.