Use Doctor Rebrand from Claude, Claude Code, or ChatGPT (MCP)

Doctor Rebrand ships a Model Context Protocol server that wraps the public API as MCP tools. Any client that speaks MCP can connect with a PAT and start calling.

Endpoint

https://mcp.doctorrebrand.com/mcp

Auth

Mint a PAT at core.doctorrebrand.com/app/settings/api-tokens and bearer it on every MCP request. The PAT's scopes decide which tools work — start with the read-only preset and add write scopes only when you trust the integration.

Claude.ai or ChatGPT (remote)

In the connector / custom-GPT setup UI, add a new MCP server:

• URL: https://mcp.doctorrebrand.com/mcp
• Auth: Bearer token = your PAT

Claude Code (local)

Add to ~/.claude/mcp.json (or a project-local .mcp.json):

{
  "mcpServers": {
    "doctor-rebrand": {
      "url": "https://mcp.doctorrebrand.com/mcp",
      "headers": {
        "Authorization": "Bearer …your-token…"
      }
    }
  }
}

What you get

The current toolset is intentionally small (8 tools — identity, wallet read, dry-run cost preview, branding, content reads). Mutation tools (sending a newsletter, unlocking a recruit) land in v2; until then use the HTTP API for those. See the full API reference.

Cost safety

For any operation that costs tokens, always use wallet_reserve_dry_run first to preview the cost, surface it to the user, and only proceed if they agree. Operations costing ≥ 10 tokens additionally require an explicit confirmed: true flag — server-side, regardless of which client is calling. This is your AI's safety belt against runaway spending.