WorkSync

🔄 WorkSync - Synchronize work activities across Harvest, Linear, and Slack. Features automated daily standups, AI-powered time tracking via Claude Desktop, and comprehensive activity reporting. Built with Node.js and TypeScript.

Installation

Run this command in your terminal to add the MCP server to Claude Code.

Run in terminal:

View docs

Command

claude mcp add --transport stdio qdhenry-worksync npm run mcp \
  --env HARVEST_ACCOUNT_ID="your_harvest_account_id" \
  --env HARVEST_ACCESS_TOKEN="your_harvest_access_token"

How to use

WorkSync exposes an MCP (Model Context Protocol) server that lets Claude Desktop add time entries to Harvest using natural language. In addition to Claude, the server coordinates with Slack for daily standups and integrates with Harvest and Linear to surface and log work across tools. The MCP configuration points Claude Desktop at the Worksync MCP server, enabling you to enter time via natural language prompts (for example, “log 2 hours for the API work”) and have those entries created in Harvest. The server uses an npm-based implementation that runs a dedicated MCP script, configured to access Harvest credentials and operate in your chosen workspace. You can additionally use the built-in standup features to visualize yesterday’s work and planned activities in Slack. The modular design also supports running components separately (e.g., standup posting, time-entry parsing) if you want to customize the workflow.

How to install

Prerequisites:

Node.js 20+ and npm installed
Access to Harvest API (account ID and token)
Optional: Slack webhook for standups

Step-by-step:

Clone the repository or download the Worksync MCP package
Install dependencies: npm install
Configure your environment variables (see below) and prepare the MCP script:
- HARVEST_ACCOUNT_ID: your Harvest account ID
- HARVEST_ACCESS_TOKEN: your Harvest API token
Start the MCP server (Node-based): npm run mcp
In Claude Desktop, add an MCP server entry pointing to this Worksync MCP server using the provided command and path, and include the required HARVEST credentials in the environment.

Environment variables can be set in a .env file or directly in your Claude Desktop configuration as shown in the MCP setup example.

Additional notes

Tips and notes:

Ensure Harvest API credentials have the necessary permissions to create time entries.
If you enable Linear integration, provide a valid Linear API key; otherwise, the server will still operate with Harvest standups.
The MCP server is designed to be timezone-aware; set TIMEZONE to match your location for accurate scheduling and logging.
For production deployments, consider containerizing with Docker and using the provided docker-compose setup.
If you encounter issues with Claude integration, verify the environment variables are correctly passed to the MCP process and that Claude Desktop is configured to target the correct mcp server name (e.g., worksync).
Enable DEBUG in your environment to get verbose logs for troubleshooting.

Related MCP Servers

n8n-workflow-builder

493

AI assistant integration for n8n workflow automation through Model Context Protocol (MCP). Connect Claude Desktop, ChatGPT, and other AI assistants to n8n for natural language workflow management.

Gitingest

136

mcp server for gitingest

ytt

MCP server to fetch YouTube transcripts

mcp-dock

A cross-platform MCP Server manager for Cursor, Claude, Windsurf, Zed & TRAE. Features one-click installation, multi-client sync, and a curated registry of Official & Smithery servers.

mcp_server_filesystem

MCP File System Server: A secure Model Context Protocol server that provides file operations for AI assistants. Enables Claude and other assistants to safely read, write, and list files in a designated project directory with robust path validation and security controls.

memcord

🧠 Privacy-first MCP server for AI memory management. Save, search & organize chat history with intelligent summarization.