Mosshatch

Use Mosshatch from Claude and other assistants

Mosshatch speaks the Model Context Protocol (MCP), so the assistant you already use can work with your names. It can check whether a name is free and what it costs, read and change DNS records, connect a name to Vercel, Resend or Neon, and suggest names to buy or renew. It runs on your own assistant plan; Mosshatch charges nothing for it.

Two addresses

Claude: claude.ai and the Claude apps

  1. Open Settings, then Connectors, and choose Add custom connector.
  2. Name it Mosshatch and paste one of the addresses above.
  3. For your account, connect and sign in. Choose what Claude may do, then confirm with your passkey.

A connector added on claude.ai also works in the Claude desktop and mobile apps.

Claude Code

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

Then type /mcp in Claude Code and choose Mosshatch to sign in. For search only, use https://mosshatch.com/mcp/search instead; it needs no sign-in.

ChatGPT

Turn on developer mode under Settings, Apps and connectors, Advanced. Then create a connector with one of the addresses and sign in when asked. Developer mode is still a beta in ChatGPT, so if its menus differ, follow ChatGPT's own help.

Cursor and other editors

Add this to ~/.cursor/mcp.json, and Cursor opens your browser to sign in:

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

Any assistant or editor that connects to remote MCP servers with sign-in works the same way.

Scripts, servers and people

For a program that cannot sign in, or a developer working on one of your names, make a token under Account, Connected apps. Tick only what it needs, on one name if you can, and send it as Authorization: Bearer followed by the token. Tokens end within 90 days.

Things to ask

The Vercel, Resend and Neon recipes need that provider connected once, on the name's Connect tab in Mosshatch.

What an assistant cannot do on its own

Report a security problem with the MCP servers to security@mosshatch.com; see the security policy.