Autotask

You can run the Autotask MCP Server to give AI assistants structured, programmable access to Kaseya Autotask PSA data and operations. It exposes a rich set of MCP-compliant endpoints and tools to manage companies, contacts, tickets, time entries, projects, notes, attachments, and more, with secure authentication, caching, and flexible transport options for desktop or remote deployments.

How to use

You use an MCP client to connect to the Autotask MCP Server. The server can operate locally via stdio or remotely over HTTP. When connected, you can search, create, update, and read various Autotask entities, and invoke specialized tools to perform common work tasks. The ID-to-name mapping feature enhances responses by resolving IDs to human-friendly names, and the server uses caching to improve performance while reducing repeated API calls.

How to install

Prerequisites: you need valid Autotask API credentials and an MCP-compatible client like Claude Desktop or Claude Code. You may also need Docker if you choose the remote/HTTP deployment path, or Node.js 18+ if you build from source.

# Option A: Install and run using a pre-packaged MCP Bundle (Claude Desktop)
# Start by obtaining the MCP Bundle file from the latest release and open it in Claude Desktop.

# Option B: Run with Docker (HTTP transport by default)
docker run -d \
  --name autotask-mcp \
  -p 8080:8080 \
  -e AUTOTASK_USERNAME="your-api-user@example.com" \
  -e AUTOTASK_SECRET="your-secret-key" \
  -e AUTOTASK_INTEGRATION_CODE="your-integration-code" \
  --restart unless-stopped \
  ghcr.io/wyre-technology/autotask-mcp:latest

# Health check for HTTP transport
curl http://localhost:8080/health

# Option C: From Source (Development)
# Build and run locally after cloning the repository
git clone https://github.com/wyre-technology/autotask-mcp.git
cd autotask-mcp
npm ci && npm run build

# Start the server with the built entry point
MCP_TRANSPORT=http node dist/index.js

Configuration and deployment options

The server supports multiple deployment modes. You can run in local stdio mode for Claude Desktop or in HTTP mode for remote deployments behind a gateway or over the network.

{
  "mcpServers": {
    "autotask": {
      "command": "node",
      "args": ["dist/entry.js"],
      "env": {
        "AUTOTASK_USERNAME": "your-user@company.com",
        "AUTOTASK_SECRET": "your-secret",
        "AUTOTASK_INTEGRATION_CODE": "your-code"
      }
    }
  }
}

Gateway and HTTP transport details

For remote deployments, you can enable HTTP transport to expose an MCP Streamable endpoint. The server typically listens on port 8080 and can be reached at a path like /mcp for MCP requests and /health for health checks.

Security and environment configuration

Store credentials in environment variables and limit Autotask API user permissions to the minimum required. Rotate credentials regularly. When deploying with a gateway, credentials may be injected via HTTP headers.

Troubleshooting and tips

Enable debug logging to diagnose slow responses or connectivity issues. If you encounter MCP client connection problems, verify environment variables, JSON configuration, and that you are using the correct entry point for your chosen transport.

Development notes

If you are developing locally, you can run tests, build, and start in development mode. Use npm scripts for common tasks like building, linting, and testing.

Available tools

autotask_search_companies

Search companies with filters

autotask_create_company

Create new company

autotask_update_company

Update existing company

autotask_search_contacts

Search contacts with filters

autotask_create_contact

Create new contact

autotask_search_tickets

Search tickets with filters

autotask_get_ticket_details

Get full ticket details by ID

autotask_create_ticket

Create new ticket

autotask_create_time_entry

Log time entry

autotask_search_time_entries

Search time entries with filters

autotask_search_billing_items

Search approved and posted billing items

autotask_get_billing_item

Get specific billing item by ID

autotask_search_billing_item_approval_levels

Search multi-level approval records for time entries

autotask_search_projects

Search projects with filters

autotask_create_project

Create new project

autotask_search_resources

Search resources (technicians/users)

autotask_get_ticket_note

Get ticket note

autotask_search_ticket_notes

Search ticket notes

autotask_create_ticket_note

Create ticket note

autotask_get_project_note

Get project note

autotask_search_project_notes

Search project notes

autotask_create_project_note

Create project note

autotask_get_company_note

Get company note

autotask_search_company_notes

Search company notes

autotask_create_company_note

Create company note

autotask_get_ticket_attachment

Get ticket attachment

autotask_search_ticket_attachments

Search ticket attachments

autotask_get_expense_report

Get expense report

autotask_search_expense_reports

Search expense reports

autotask_create_expense_report

Create expense report

autotask_get_quote

Get quote

autotask_search_quotes

Search quotes

autotask_create_quote

Create quote

autotask_search_invoices

Search invoices

autotask_search_contracts

Search contracts

autotask_search_configuration_items

Search configuration items (assets)

autotask_search_tasks

Search project tasks

autotask_create_task

Create project task

autotask_test_connection

Test API connectivity