Harvest Time Tracking MCP Server
LocalAbout Harvest Time Tracking MCP Server
Harvest Time Tracking MCP Server lets LLM clients access Harvest account data and manage time entries. Its tools cover company, client, project, user, task assignments, and time-entry lookup and creation.
Key features
- 01Retrieve company information
- 02Search and inspect clients and projects
- 03List users and project assignments
- 04Search and retrieve time entries
- 05Create time entries
Use cases
- 01Review client and project details
- 02Check project users and task assignments
- 03Find time entries by date or account filters
- 04Record hours against a project task
MCP support
Exposes Harvest account operations as MCP tools, including lookups for company, clients, projects, users, and time entries, plus time-entry creation.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y @harvest/mcp-server-typescriptHARVEST_ACCOUNT_IDHARVEST_ACCESS_TOKENSecretHARVEST_BASE_URLLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEHarvest Time Tracking MCP Server
A TypeScript based MCP (Model Context Protocol) server for Harvest time tracking app enabling LLM clients to interact with Harvest account.
Installation
npx -y @mayankg0324/mcp-server-harvest
Prerequisites
- Node.js >= 18 (or Bun Runtime for development)
- Harvest account with API access
- API Personal Access Token from Harvest
- Account ID from Harvest
More details on how to create a Personal Access Token can be found in the Harvest API documentation.
Available Tools
| Tool | Description | Inputs | Output |
|---|---|---|---|
get-company | Get company information | None | Company Information (Name, URL, ID and Currency) |
search-clients | Searches for clients | • name: string (optional) - Client name • isActive: boolean (optional) - Active status | List of clients with IDs and details |
get-client | Get client information | • clientId: string - Client ID | Detailed client information |
search-projects | Searches for projects | • name: string (optional) - Project name • clientId: string (optional) - Client ID • isActive: boolean (optional) - Active status | List of projects with IDs and details |
get-project | Get project information | • projectId: string - Project ID | Detailed project information |
list-users | Lists all users | None | List of users with IDs and details |
get-user | Get user information | • userId: string - User ID | Detailed user information |
list-project-users | List project user assignments | • projectId: string - Project ID | List of users assigned to project |
list-project-tasks | List project task assignments | • projectId: string - Project ID | List of tasks assigned to project |
search-time-entries | Searches for time entries | • from: string (optional) - Start date (YYYY-MM-DD) • to: string (optional) - End date (YYYY-MM-DD) • userId: string (optional) - User ID • projectId: string (optional) - Project ID • clientId: string (optional) - Client ID | List of time entries with details |
get-time-entry | Get time entry details | • timeEntryId: string - Time Entry ID | Detailed time entry information |
create-time-entry | Create a new time entry | • projectId: string - Project ID • taskId: string - Task ID • userId: string - User ID • hours: number - Hours spent • notes: string (optional) - Additional notes | Created time entry details |
Note: More tools coming soon.
Usage
Claude Desktop
Open the Claude desktop config file claude-desktop-config.json and add the following under mcpServers:
{
"mcpServers": {
"Harvest": {
"command": "npx",
"args": [
"-y",
"@mayankg0324/mcp-server-harvest"
],
"env": {
"HARVEST_ACCOUNT_ID": "<Harvest Account ID>",
"HARVEST_ACCESS_TOKEN": "<Harvest Personal Access Token>",
"HARVEST_BASE_URL": "<Harvest Base URL>"
}
}
}
}
Cursor IDE
Open the Cursor IDE config file ~/.cursor/mcp.json or <project-root>/.cursor/mcp.json and add the following under mcpServers:
{
"mcpServers": {
"Harvest": {
"command": "npx",
"args": [
"-y",
"@mayankg0324/mcp-server-harvest"
],
"env": {
"HARVEST_ACCOUNT_ID": "<Harvest Account ID>",
"HARVEST_ACCESS_TOKEN": "<Harvest Personal Access Token>",
"HARVEST_BASE_URL": "<Harvest Base URL>"
}
}
}
}
Troubleshooting
spawn npx ENOENT or env: node: No such file or directory
This happens when your MCP client (Claude Desktop, Cursor) cannot find npx or node in its PATH. GUI applications often don't inherit your shell's PATH, especially if you use a version manager like nvm, fnm, or volta.
Fix: Use the full path to npx and include your Node bin directory in the PATH env variable:
# Find your npx and node paths
which npx
# e.g. /Users/you/.nvm/versions/node/v22.20.0/bin/npx
Then update your MCP config:
{
"mcpServers": {
"Harvest": {
"command": "/Users/you/.nvm/versions/node/v22.20.0/bin/npx",
"args": [
"-y",
"@mayankg0324/mcp-server-harvest"
],
"env": {
"PATH": "/Users/you/.nvm/versions/node/v22.20.0/bin:/usr/bin:/bin",
"HARVEST_ACCOUNT_ID": "<Harvest Account ID>",
"HARVEST_ACCESS_TOKEN": "<Harvest Personal Access Token>",
"HARVEST_BASE_URL": "<Harvest Base URL>"
}
}
}
}
Local Development
-
Install dependencies:
npm install -
Build the project:
npm run buildThis will create a
distdirectory containing the compiled JavaScript files. -
To use the local development version, modify your MCP server config to point to the local build:
{ "mcpServers": { "Harvest": { "command": "node", "args": [ "path/to/local/build/dist/index.js" ], "env": { "HARVEST_ACCOUNT_ID": "<Harvest Account ID>", "HARVEST_ACCESS_TOKEN": "<Harvest Personal Access Token>", "HARVEST_BASE_URL": "<Harvest Base URL>" } } } } -
Restart the client application (Claude, Cursor) to apply the changes.
License
MIT
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.