Harvest MCP server

This MCP server enables natural language time tracking in Harvest via Claude. It understands plain language inputs for logging time entries, handles leave requests automatically, and provides detailed time reports with intelligent date parsing.

Installation

Prerequisites

• Node.js installed
• A Harvest account
• Personal access token from Harvest Developer Tools
• Account ID (shown on the same page as your token)

Setup Steps

1. Install the Claude desktop app

2. Clone the repository:

   git clone https://github.com/adrian-dotco/harvest-mcp-server.git
   cd harvest-mcp-server
   

3. Install dependencies and build:

   npm install
   npm run build
   

4. Run the setup script:

   node build/setup.js
   

5. Follow the prompts to enter your:

   • Harvest Personal Access Token
   • Harvest Account ID
   • Standard work day hours (default: 7.5)
   • Timezone (default: Australia/Perth)

6. Restart the Claude desktop app

Updating the Server

To update to the latest version:

git pull
npm install
npm run build

Using the MCP Server

The server provides several tools for interacting with Harvest through natural language commands.

Logging Time Entries

Use the log_time function with natural language descriptions:

Regular time entries:

"2 hours on Project X doing development work today"
"45 minutes on Project Y testing yesterday"
"3.5 hours on Project Z meetings last Friday"

Leave requests:

"I'm off sick today"
"I'm unwell today"
"Taking annual leave next week"

Getting Time Reports

Use the get_time_report function with various natural language queries:

Time period options:

"Show time report for last month"
"Get time summary for this week"
"Show hours from January 1st to January 31st"
"Report time for Q1"
"Show me yesterday's hours"

Report types:

• By Project (default):

  "Show time report for last month"
  "Get project hours for this week"
  

• By Client:

  "Show time report by client for this month"
  "Get hours by client for Q1"
  

• By Task:

  "Show time summary by task for January"
  "Get task breakdown for last week"
  

• By Team Member:

  "Show team hours for last week"
  "Get time report by user for this month"
  

Listing Projects

View all available Harvest projects:

List my projects

Listing Tasks

View available tasks for a specific project:

Show tasks for Project X

Viewing Recent Entries

Check your recent time entries:

Show my recent time entries

Configuration Options

The server supports these environment variables:

• HARVEST_ACCESS_TOKEN: Your Harvest personal access token
• HARVEST_ACCOUNT_ID: Your Harvest account ID
• STANDARD_WORK_DAY_HOURS: Default hours for a full work day (default: 7.5)
• TIMEZONE: Your timezone (default: Australia/Perth)

These can be set either through the setup script or manually in your environment.

How to add this MCP server to Cursor

There are two ways to add an MCP server to Cursor. The most common way is to add the server globally in the ~/.cursor/mcp.json file so that it is available in all of your projects.

If you only need the server in a single project, you can add it to the project instead by creating or adding it to the .cursor/mcp.json file.

Adding an MCP server to Cursor globally

To add a global MCP server go to Cursor Settings > MCP and click "Add new global MCP server".

When you click that button the ~/.cursor/mcp.json file will be opened and you can add your server like this:

{
    "mcpServers": {
        "cursor-rules-mcp": {
            "command": "npx",
            "args": [
                "-y",
                "cursor-rules-mcp"
            ]
        }
    }
}

Adding an MCP server to a project

To add an MCP server to a project you can create a new .cursor/mcp.json file or add it to the existing one. This will look exactly the same as the global MCP server example above.

How to use the MCP server

Once the server is installed, you might need to head back to Settings > MCP and click the refresh button.

The Cursor agent will then be able to see the available tools the added MCP server has available and will call them when it needs to.

You can also explictly ask the agent to use the tool by mentioning the tool name and describing what the function does.