Lazy Toggl MCP
Simple unofficial MCP server to track time via Toggl API
Lazy Toggl MCP Server
A Model Context Protocol (MCP) server that provides tools for interacting with Toggl time tracking.
Features
- Start/stop time tracking
- Get current entry
- List workspaces
API
Tools
-
start_tracking
- Start tracking time for a new task
- Inputs:
title(string): Title/description of the task to trackworkspace_id(integer): Workspace ID (optional, uses default if not provided)project_id(integer): Project ID (optional)tags(string[]): List of tags (optional)
-
stop_tracking
- Stop the currently running time entry
- No input required
- Returns confirmation of stopped time entry
-
list_workspaces
- List all available workspaces
- No input required
- Returns list of workspaces with their IDs and names
-
show_current_time_entry
- Show the currently running time entry, if any
- No input required
- Returns:
- If tracking: Task description, entry ID, workspace, start time, running duration, tags, and project (if any)
- If not tracking: A message indicating no time entry is currently running
Integration with Toggl Track API
This server uses the Toggl Track API v9. The following endpoints are utilized:
GET /me- Get user informationGET /workspaces- List workspacesGET /me/time_entries/current- Get current running time entryPOST /workspaces/{workspace_id}/time_entries- Start time trackingPATCH /workspaces/{workspace_id}/time_entries/{time_entry_id}/stop- Stop time tracking
Installation
- Clone/create this project
- Install dependencies with
uv:cd lazy-toggl-mcp uv sync
Configuration
Get Your Toggl API Token
- Go to Toggl Track
- Sign in to your account
- Click on your profile picture/avatar in the top right corner
- Go to "Profile" or "Settings"
- Find your "API Token" - copy this long string of characters
Configure MCP Server
Add the following configuration to your MCP settings file:
{
"mcpServers": {
"lazy-toggl-mcp": {
"autoApprove": [],
"disabled": false,
"timeout": 60,
"type": "stdio",
"transportType": "stdio",
"command": "uv",
"args": [
"run",
"--directory",
"/path/to/lazy-toggl-mcp",
"python",
"server.py"
],
"env": {
"TOGGL_API_TOKEN": "your-actual-api-token-here"
}
}
}
}
Important: Replace /path/to/lazy-toggl-mcp with the actual path to this project and your-actual-api-token-here with your real Toggl API token.
Project Structure
lazy-toggl-mcp/
├── src/
│ └── toggl_server/
│ ├── __init__.py # Package initialization
│ ├── main.py # MCP server implementation (new structure)
│ ├── models.py # Data models and type definitions
│ ├── toggl_api.py # Toggl API client
│ └── utils.py # Utility functions
├── main.py # CLI interface for testing
├── server.py # Main MCP server entry point
├── pyproject.toml # Project configuration and dependencies
├── README.md # This file
├── uv.lock # Dependency lock file
├── .gitignore # Git ignore patterns
└── .python-version # Python version specification
License
MIT License - feel free to modify and use as needed.
相關伺服器
Kone.vc
贊助Monetize your AI agent with contextual product recommendations
Nextcloud Calendar
CalDAV Nectcloud calendar integration. Manage calendars, events, attendees, etc.
Docmost
An open-source collaborative wiki and documentation software with AI integration via MCP.
clipboard-mcp
MCP server that reads and writes the system clipboard — tables, text, code, JSON, URLs, images, and more. Preserves spreadsheet structure (rows/columns) that is lost when pasting into Claude directly. Claude can also write results back to your clipboard.
Smartsheet
Integrate with Smartsheet for project management and data analytics, requiring an API access token.
Avocado AI
Collaborative AI creative workspace for agencies and ecommerce teams to generate on-brand images, videos, and ad creative at scale.
Blender AI MCP
Modular MCP Server + Blender Addon for AI-Driven 3D Modeling.
UpTier
Desktop task manager with clean To Do-style UI and 25+ MCP tools for prioritization, goal tracking, and multi-profile workflows.
amem
The memory layer for AI coding tools. Local-first, semantic, 9 MCP tools with consolidation and project scoping. Works with Claude Code, Cursor, Windsurf & any MCP client.
Hjarni
A Markdown-based second brain with a hosted MCP server. Your AI reads, searches, creates, and links your notes. Free tier included.
screenpipe
use 24/7 desktop memory as context in AI