Skip to main content
Glama

VeryCloud MCP

Manage a VeryCloud/Wisp game server from ChatGPT, Claude, Cursor, or another MCP client.

Supports server status, files and logs, console commands, power controls, audit logs, and backups. Your VeryCloud API token stays on the MCP host.

Install

Requires Python 3.11+ and a VeryCloud API token.

git clone https://github.com/BreezeDelegate/verycloud-mcp.git
cd verycloud-mcp
python -m venv .venv

Windows PowerShell:

.\.venv\Scripts\Activate.ps1
python -m pip install -e .
Copy-Item .env.example .env

Linux/macOS:

source .venv/bin/activate
python -m pip install -e .
cp .env.example .env

Set these two values in .env:

VERYCLOUD_API_TOKEN=your_verycloud_api_token
VERYCLOUD_SERVER_ID=your_server_id

Then:

verycloud-mcp --check
verycloud-mcp

Local MCP URL: http://127.0.0.1:8000/mcp

Related MCP server: Simple SSH MCP Server

Connect

ChatGPT: private/local test with Secure MCP Tunnel

This is the recommended first real-world test because the MCP server remains bound to localhost and does not need a public inbound port.

  1. In OpenAI Platform, create a Secure MCP Tunnel and a separate runtime API key with the required tunnel permissions.

  2. Put the values in your local .env only:

CONTROL_PLANE_TUNNEL_ID=tunnel_...
CONTROL_PLANE_API_KEY=your_runtime_api_key
  1. Install the official OpenAI tunnel-client and ensure tunnel-client is available in your PATH.

  2. Start VeryCloud MCP and the tunnel together:

Windows PowerShell:

.\scripts\run-chatgpt-tunnel.ps1

Linux/macOS:

chmod +x scripts/run-chatgpt-tunnel.sh
./scripts/run-chatgpt-tunnel.sh
  1. In ChatGPT Developer Mode, create a custom app, choose Tunnel as the connection type, select the tunnel, scan the tools, and test the draft app.

Official OpenAI tunnel documentation: https://developers.openai.com/api/docs/guides/secure-mcp-tunnels

Never paste the VeryCloud API token or OpenAI runtime API key into chat, issues, commits, or screenshots.

Claude Code

claude mcp add --transport http verycloud http://127.0.0.1:8000/mcp

For a remote instance protected with VERYCLOUD_MCP_AUTH_TOKEN:

claude mcp add --transport http verycloud https://mcp.example.com/mcp --header "Authorization: Bearer your_mcp_token"

Cursor

Add the MCP URL as a Streamable HTTP server in Cursor's MCP settings. Project configs live in .cursor/mcp.json; global configs live in ~/.cursor/mcp.json.

Allow changes

Write actions are off by default. Enable only what you want in .env:

VERYCLOUD_ALLOW_COMMANDS=true
VERYCLOUD_ALLOW_FILE_WRITES=true
VERYCLOUD_ALLOW_POWER=true
VERYCLOUD_ALLOW_BACKUPS=true
VERYCLOUD_ALLOW_DESTRUCTIVE=false

VERYCLOUD_ALLOW_DESTRUCTIVE=true is additionally required for file deletion, backup restore/delete, and force-kill.

Remote / Docker

For a remote listener:

VERYCLOUD_MCP_HOST=0.0.0.0
VERYCLOUD_MCP_AUTH_TOKEN=use_a_long_random_token_here
VERYCLOUD_MCP_ALLOWED_HOSTS=mcp.example.com,mcp.example.com:443

Put HTTPS in front of the MCP server. VERYCLOUD_MCP_AUTH_TOKEN protects the MCP endpoint; it is not your VeryCloud token.

docker build -t verycloud-mcp .
docker run --rm -p 8000:8000 --env-file .env -e VERYCLOUD_MCP_HOST=0.0.0.0 verycloud-mcp

Never commit .env or any token. Non-commercial use only; see LICENSE. Contributions: CONTRIBUTING.md.

Maintenance

ActivityMaintained
ResponsivenessSyncing

Related MCP Connectors

Related MCP Servers

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/BreezeDelegate/verycloud-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server