Connect

Every client connects to the same URL over Streamable HTTP:

https://sleepytotem.xyz/mcp

/sse points to the same handler for clients that expect that path.

Claude (Anthropic)

Claude Code

claude mcp add --transport http mak-suite https://sleepytotem.xyz/mcp

Claude.ai and Claude Desktop

Open connectors

Go to Customize → Connectors and click +. On Team and Enterprise plans, an owner adds it first under Organization settings → Connectors.

Add a custom connector

Choose Add custom connector, name it mak-suite, and enter the URL above. Leave OAuth empty.

Enable it in a chat

Turn on mak-suite from the tools menu in the composer.

Claude API

The MCP connector needs both mcp_servers and a matching mcp_toolset.

curl https://api.anthropic.com/v1/messages \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "anthropic-beta: mcp-client-2025-11-20" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-opus-5",
    "max_tokens": 16000,
    "mcp_servers": [
      { "type": "url", "name": "mak-suite", "url": "https://sleepytotem.xyz/mcp" }
    ],
    "tools": [{ "type": "mcp_toolset", "mcp_server_name": "mak-suite" }],
    "messages": [{ "role": "user", "content": "What is the MAK_API frame layout?" }]
  }'

ChatGPT and Codex (OpenAI)

ChatGPT

Available on Plus, Pro, Business, Enterprise and Edu on the web.

Enable developer mode

Go to Settings → Security and login and turn on Developer mode.

Create an app

Add a new developer-mode app with the URL above and No authentication.

Use it

Pick mak-suite from the developer mode tools in the composer.

Codex CLI, IDE extension and desktop app

codex mcp add mak-suite --url https://sleepytotem.xyz/mcp

Or add it to ~/.codex/config.toml. The Codex CLI, IDE extension and desktop app all read this file.

[mcp_servers.mak-suite]
url = "https://sleepytotem.xyz/mcp"

OpenAI API

curl https://api.openai.com/v1/responses \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "gpt-6-astra",
    "tools": [{
      "type": "mcp",
      "server_label": "mak-suite",
      "server_description": "MAKCU and MAKXD protocol docs and SDK source.",
      "server_url": "https://sleepytotem.xyz/mcp",
      "require_approval": "never"
    }],
    "input": "What is the MAK_API frame layout?"
  }'

Gemini (Google)

Gemini CLI

gemini mcp add --transport http mak-suite https://sleepytotem.xyz/mcp

Or add it to ~/.gemini/settings.json:

{
  "mcpServers": {
    "mak-suite": { "httpUrl": "https://sleepytotem.xyz/mcp" }
  }
}

GitHub Copilot

VS Code

Run MCP: Open User Configuration from the Command Palette, or create .vscode/mcp.json in a workspace:

{
  "servers": {
    "mak-suite": {
      "type": "http",
      "url": "https://sleepytotem.xyz/mcp"
    }
  }
}

Copilot CLI

copilot mcp add --transport http mak-suite https://sleepytotem.xyz/mcp

Le Chat (Mistral)

Open connectors

Go to Intelligence → Connectors and click + Add Connector.

Add a custom MCP connector

On the Custom MCP Connector tab, name it mak-suite, enter the URL above and click Connect.

Choose authentication

Select No authentication.

Editors and agents

These work with any model the editor is set up to use.

{
  "mcpServers": {
    "mak-suite": { "url": "https://sleepytotem.xyz/mcp" }
  }
}

{
  "mcpServers": {
    "mak-suite": { "serverUrl": "https://sleepytotem.xyz/mcp" }
  }
}

{
  "mcpServers": {
    "mak-suite": { "url": "https://sleepytotem.xyz/mcp" }
  }
}

{
  "context_servers": {
    "mak-suite": { "url": "https://sleepytotem.xyz/mcp" }
  }
}

{
  "mcpServers": {
    "mak-suite": {
      "type": "streamableHttp",
      "url": "https://sleepytotem.xyz/mcp"
    }
  }
}

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "mak-suite": {
      "type": "remote",
      "url": "https://sleepytotem.xyz/mcp",
      "enabled": true
    }
  }
}

For JetBrains, paste the snippet under Settings → Tools → AI Assistant → Model Context Protocol (MCP).

Local models

LM Studio

In the right sidebar, open Program → Install → Edit mcp.json:

{
  "mcpServers": {
    "mak-suite": { "url": "https://sleepytotem.xyz/mcp" }
  }
}

The model you load needs to support tool use.

Clients that only speak stdio

Bridge through mcp-remote:

{
  "mcpServers": {
    "mak-suite": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://sleepytotem.xyz/mcp"]
    }
  }
}

HTTP endpoints

Path	Method	Returns
`/mcp`	`POST`	JSON-RPC 2.0, one message or a batch. `/mcp/`, `/sse` and `/` answer the same way.
`/mcp`	`DELETE`	`204`. Sessions only record which firmware notices were sent, so there is nothing to close.
`/mcp`	`GET`	MCP clients (`Accept: text/event-stream` or `application/json`) get `405`: the server keeps no state, so it has no stream to open. Browsers are redirected to this documentation.
`/mcp/health`	`GET`	Server version, snapshot commit, when upstream was last checked, file count, size and tool names. Add `?refresh=1` to check upstream now.
`/.well-known/mcp.json`	`GET`	A machine-readable description for directories and install tools. Also served at `/mcp/.well-known/mcp.json`.

Get started

Reference

Claude (Anthropic)

Claude Code

Claude.ai and Claude Desktop

Claude API

ChatGPT and Codex (OpenAI)

ChatGPT

Codex CLI, IDE extension and desktop app

OpenAI API

Gemini (Google)

Gemini CLI

GitHub Copilot

VS Code

Copilot CLI

Le Chat (Mistral)

Editors and agents

Local models

LM Studio

Clients that only speak stdio

HTTP endpoints

​Claude (Anthropic)

​Claude Code

​Claude.ai and Claude Desktop

​Claude API

​ChatGPT and Codex (OpenAI)

​ChatGPT

​Codex CLI, IDE extension and desktop app

​OpenAI API

​Gemini (Google)

​Gemini CLI

​GitHub Copilot

​VS Code

​Copilot CLI

​Le Chat (Mistral)

​Editors and agents

​Local models

​LM Studio

​Clients that only speak stdio

​HTTP endpoints

Claude (Anthropic)

Claude Code

Claude.ai and Claude Desktop

Claude API

ChatGPT and Codex (OpenAI)

ChatGPT

Codex CLI, IDE extension and desktop app

OpenAI API

Gemini (Google)

Gemini CLI

GitHub Copilot

VS Code

Copilot CLI

Le Chat (Mistral)

Editors and agents

Local models

LM Studio

Clients that only speak stdio

HTTP endpoints