# Trainzilla Coach MCP

> Trainzilla Coach MCP is an open-source MCP server for productivity. MCP server letting AI clients manage coaching clients, plans, sessions and billing. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** API, CLI, Self-hosted
- **Repository:** https://github.com/trainzilla/tzilla-mcp
- **MCP Registry name:** io.github.Trainzilla/trainzilla-mcp
- **Categories:** [productivity](https://ai-powered.com/categories/productivity)
- **Listing updated:** 2026-11-19

## About

The README describes this as a Model Context Protocol server that sits in front of Trainzilla's existing GraphQL API, so an AI client such as Claude Desktop, Claude Code or Cursor can act on a coach's account. Access goes through the backend, which keeps its own authentication and business rules in force rather than giving the client direct database access.

The tool set is large and covers read operations on clients, habits, workout and diet plans, check-ins, sessions and subscriptions, plus offline calculators for energy needs, macros and one-rep max. Write actions such as creating habits, check-ins, sessions and plans return a preview unless the caller sets an explicit confirm flag. The registry entry advertises a remote streamable-HTTP endpoint with one-click OAuth.

The repository is aimed at fitness and nutrition coaches who run their practice on Trainzilla and want an assistant to review client progress or handle routine admin. The README marks the local build as an MVP, and its roadmap lists full OAuth 2.1 and a hosted endpoint as still unfinished.

## Features

- **Client and habit reads:** Lists clients, client profiles, habit compliance and recent habit activity from the coach's account.
- **Confirm-gated writes:** Create and assign actions return a preview and only change data when confirm is set to true.
- **Offline calculators:** Computes energy expenditure, macro splits by strategy and one-rep-max estimates without network calls.
- **Billing and session overview:** Reads subscriptions, billing summaries and scheduled sessions for a coach.
- **Client profile resource:** Exposes a client's profile as a JSON resource that an AI client can load.
- **Weekly review prompt:** Includes a prompt that gathers profile, habit, compliance and session data into a read-only weekly review.

## Good for

- Reviewing a coach's client roster and spotting clients who need attention
- Drafting and assigning workout or diet plans through an AI assistant
- Checking habit compliance and session history before a client call
- Scheduling sessions and check-ins while previewing changes first
- Moving client records onto Trainzilla from a file with a bulk invite tool

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http trainzilla-mcp https://api.tzilla.live/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://api.tzilla.live/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "trainzilla-mcp": {
      "url": "https://api.tzilla.live/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "trainzilla-mcp": {
      "type": "http",
      "url": "https://api.tzilla.live/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "trainzilla-mcp": {
      "serverUrl": "https://api.tzilla.live/mcp"
    }
  }
}
```

### Grok Bot (Grok Bot desktop app → Connect apps → Add a custom connector)

```sh
https://api.tzilla.live/mcp
```

Enter a Name (for example "Trainzilla Coach MCP"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "trainzilla-mcp": {
      "type": "http",
      "url": "https://api.tzilla.live/mcp"
    }
  }
}
```

## Tools

- `whoami`: Returns the identity of the signed-in coach.
- `list_clients`: Lists the coach's clients.
- `get_client_profile`: Returns a single client's profile.
- `list_client_habits`: Lists habits assigned to a client.
- `get_habit_compliance`: Returns habit compliance for a client.
- `list_sessions`: Lists coaching sessions.
- `billing_summary`: Returns a billing summary for the coach.
- `calc_tdee`: Calculates BMR, TDEE and recommended calories offline.
- `calc_macros`: Calculates macro split by strategy offline.
- `calc_1rm`: Estimates one-rep max and percentage-based weights offline.
- `create_habit`: Creates a habit, returning a preview unless confirmed.
- `schedule_session`: Schedules a session, returning a preview unless confirmed.

## FAQ

### Does the server store coach credentials?

In HTTP mode the README says the server never stores tokens and forwards each caller's API key to the GraphQL API. Local stdio mode reads access and refresh tokens from environment variables.

### Can Trainzilla Coach MCP process refunds or payments?

No. The README says taking or refunding a payment is not available because the backend has no coach-facing refund operation.

### Is the hosted endpoint live for public use?

The README describes the remote HTTP transport as localhost-only for now, and lists hosting with TLS and full OAuth 2.1 as roadmap items still to do.

### Is Trainzilla Coach MCP free?

Yes. Trainzilla Coach MCP is open source, so you can use and self-host it for free.

### How do I add Trainzilla Coach MCP to Claude Code?

Run this command in your terminal: claude mcp add --transport http trainzilla-mcp https://api.tzilla.live/mcp

### What are the best alternatives to Trainzilla Coach MCP?

The closest alternatives to Trainzilla Coach MCP are Airtable MCP, gitscrum-core and upcoach.

## Where this comes from

- Description: written from the project's README and its MCP Registry entry.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [Airtable MCP](https://ai-powered.com/mcp/airtable-mcp-2): Airtable MCP is an open-source MCP server for databases.
- [gitscrum-core](https://ai-powered.com/mcp/gitscrum-core): gitscrum-core is an MCP server for productivity.
- [upcoach](https://ai-powered.com/mcp/upcoach): upcoach is an MCP server for education and exams.

---
Source: https://ai-powered.com/mcp/trainzilla-coach-mcp
