Model Context Protocol (MCP)

The Wavedash MCP server gives AI agents current Wavedash knowledge. Agents can search the docs, plan an SDK integration for your engine, validate wavedash.toml, and walk through upload and release. Wavedash hosts the server, so there is nothing to install or run locally.

Server details

• URL: https://mcp.wavedash.com/mcp
• Transport: Streamable HTTP
• Authentication: none

The server only serves public documentation, so there is no sign-in, API key, or OAuth step. It can't access Wavedash accounts, private projects, builds, local files, or API keys.

Setup

Every client uses the same URL. Find your agent below.

Claude Code

Run this command in your project:

claude mcp add --transport http wavedash https://mcp.wavedash.com/mcp

This adds the server for the current project on your machine. Add --scope user to make it available in every project, or --scope project to write it to .mcp.json so teammates get it too.

Run claude mcp list to check the connection, or /mcp inside a session.

Claude (web and desktop)

Add Wavedash as a custom connector on claude.ai or in Claude Desktop:

1. Go to Customize > Connectors.
2. Click + Add, then Add custom connector.
3. Name it Wavedash and enter https://mcp.wavedash.com/mcp as the URL.
4. Under Authentication, choose No sign in, then click Add.

On Team and Enterprise plans, an owner adds the connector under Organization settings > Connectors (Add > Custom > Web). Each member then clicks Connect on it in Customize > Connectors.

To use it in a conversation, click + in the chat box, select Connectors, and turn on Wavedash.

Codex

The Codex CLI and IDE extension share configuration, so one command sets up both:

codex mcp add wavedash --url https://mcp.wavedash.com/mcp

Or add the server to ~/.codex/config.toml yourself:

[mcp_servers.wavedash]
url = "https://mcp.wavedash.com/mcp"

Run codex mcp list to confirm the server is enabled.

Cursor

Add Wavedash to Cursor with one click, or add it to .cursor/mcp.json in your project (~/.cursor/mcp.json for every project):

{
  "mcpServers": {
    "wavedash": {
      "url": "https://mcp.wavedash.com/mcp"
    }
  }
}

The Cursor CLI reads the same files. Approve the server once with cursor-agent mcp enable wavedash.

VS Code

Add the server to .vscode/mcp.json in your workspace, which also shares it with your team:

{
  "servers": {
    "wavedash": {
      "type": "http",
      "url": "https://mcp.wavedash.com/mcp"
    }
  }
}

Or run MCP: Add Server from the Command Palette, pick the HTTP server type, and enter the URL. From a terminal, this command adds it to your user profile:

code --add-mcp '{"name":"wavedash","type":"http","url":"https://mcp.wavedash.com/mcp"}'

The Wavedash tools are available in Copilot Chat's agent mode.

Gemini CLI

Run this command in your project:

gemini mcp add --transport http wavedash https://mcp.wavedash.com/mcp

This writes to .gemini/settings.json in the current project. Add --scope user to write to ~/.gemini/settings.json instead.

If you edit the file by hand, use httpUrl. Gemini CLI reads url as an SSE endpoint, which this server doesn't provide.

{
  "mcpServers": {
    "wavedash": {
      "httpUrl": "https://mcp.wavedash.com/mcp"
    }
  }
}

Zed

Open Settings > AI > MCP Servers, click Add Server, choose Add Remote Server, and enter the URL. Or add it to your settings file (zed: open settings file):

{
  "context_servers": {
    "wavedash": {
      "url": "https://mcp.wavedash.com/mcp"
    }
  }
}

A green dot next to the server on the MCP Servers page means it's running.

opencode

Add the server to opencode.json in your project root, or to ~/.config/opencode/opencode.json for every project:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "wavedash": {
      "type": "remote",
      "url": "https://mcp.wavedash.com/mcp",
      "enabled": true
    }
  }
}

Run opencode mcp list to check the connection.

Other clients

Any client that supports remote MCP servers over Streamable HTTP can connect with https://mcp.wavedash.com/mcp and no authentication.

If your client only supports local (stdio) servers, bridge to the remote server with mcp-remote. It requires Node.js.

{
  "mcpServers": {
    "wavedash": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://mcp.wavedash.com/mcp"]
    }
  }
}

Try it

Once the server is connected, ask your agent something like:

• "Use the Wavedash MCP to plan adding achievements and cloud saves to my Godot game."
• "Set up the Wavedash CLI and SDK for my Phaser project."
• "Check this wavedash.toml for problems." (Paste the file.)
• "What do I need to do before uploading my first build to Wavedash?"

For build, integration, and deploy requests, the server tells agents to start with wavedash_implementation_planner.

Tools

The server exposes four tools:

Engine, topic, and feature pages are resolved from this site's search index at runtime, so new docs pages are available to agents without a server change.

Authenticated steps

For authenticated actions, the MCP server tells the human which Wavedash CLI or Developer Portal step to perform. It cannot sign in, create projects, create API keys, upload builds, publish builds, or read local files.

Use the Wavedash CLI for local project actions such as authentication, initialization, local testing, build upload, and publish.

Troubleshooting

• No tools appear. Check that the URL ends in /mcp and that the client is using Streamable HTTP. https://mcp.wavedash.com/ without /mcp is an info endpoint that returns the server version and tool list as JSON. It's useful for checking that the server is reachable, but clients can't connect to it.
• The client asks how to sign in. No sign-in is needed. Choose the client's no-authentication option, such as No sign in in Claude.
• The client only supports SSE. The server doesn't have an SSE endpoint. If the client also supports local (stdio) servers, use the mcp-remote bridge.

See also

• Agents and AI
• Agent skills
• Wavedash CLI