Skip to main content
Glama
emre6943

Trackfusion MCP Server

by emre6943
README.md
# Trackfusion MCP Server

MCP (Model Context Protocol) server for Trackfusion. Exposes projects & tasks to AI assistants like Claude, OpenClaw, etc.

## Setup

```bash
npm install
npm run build
```

## Configuration

Copy the example env file and fill in your API key:

```bash
cp .env.example .env
```

Edit `.env` with your values:

```
TRACKFUSION_API_KEY=tf_your_key_here
TRACKFUSION_API_URL=https://europe-west1-oz-track.cloudfunctions.net/api  # optional, defaults to production
```

The server loads `.env` automatically via `dotenv`.

## Usage

### With OpenClaw (mcporter)

```bash
mcporter add trackfusion --stdio "node /path/to/trackfusion/mcp/dist/index.js" --env TRACKFUSION_API_KEY=tf_xxx
```

### With Claude Desktop

Add to `claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "trackfusion": {
      "command": "node",
      "args": ["/path/to/trackfusion/mcp/dist/index.js"],
      "env": {
        "TRACKFUSION_API_KEY": "tf_your_key_here"
      }
    }
  }
}
```

## Tools

| Tool | Description |
|------|-------------|
| `list_projects` | List all projects with task counts |
| `list_tasks` | List tasks in a project (optional status filter) |
| `get_task` | Get full task details |
| `create_task` | Create a new task |
| `update_task` | Update a task (status, title, priority, etc.) |

## Getting an API Key

1. Log in to Trackfusion
2. Go to Settings → API Keys
3. Click "New Key", select scopes, and copy the generated key

TDQS

C2.9/5.0

Scored across 76 tools

Disambiguation3/5

Most tools are grouped by clear noun objects, but the sheer number and near-synonymous actions create ambiguity: search_foods vs lookup_barcode vs search_community_foods, get_day_entry vs journal entries, create_item vs create_food_definition, and update_* vs set_* vs add_* all blur boundaries. Descriptions usually clarify, but an agent selecting among 76 tools will still struggle to pick the right one.

Naming Consistency4/5

Names overwhelmingly follow a verb_noun snake_case pattern (list/create/update/delete/get), which is predictable and readable. Consistency is weakened by irregular singular/plural forms (spending/spendings, income/incomes, person/people) and by interchangeable verbs such as add_interaction vs create_*, set_nutrition_goals vs update_*, and log_meal_template vs create_meal_entry.

Tool Count1/5

With 76 tools, the server presents an extreme, unwieldy flat namespace that is very difficult for an agent to navigate during tool selection. Even though each tool spans a different module, the aggregate count far exceeds what is appropriate for an MCP server, especially with many single-purpose tools like water intake helpers.

Completeness2/5

The server covers many domains, but several core lifecycles are incomplete: projects can be listed/shared but not created, tasks have no delete, income has no update/delete, workouts have no update/delete, and journal/day entries cannot be edited or deleted. These gaps will cause agent failures when users try to correct or manage data beyond basic creation.

Maintenance

ActivityInactive
ResponsivenessNo issues