playbooks

home / mcp / jobnimbus mcp server

JobNimbus MCP Server

JobNimbus MCP Remote Server - Secure remote access to JobNimbus from any Claude Desktop client. 48 tools, zero-storage security, auto-deploy.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "benitocabrerar-jobnimbus-mcp-remote": {
      "command": "node",
      "args": [
        "C:/ruta/a/examples/mcp-client.js"
      ],
      "env": {
        "MCP_SERVER_URL": "https://tu-servidor.onrender.com",
        "JOBNIMBUS_API_KEY": "tu_api_key_stamford",
        "JOBNIMBUS_INSTANCE": "stamford"
      }
    }
  }
}

You can securely connect Claude Desktop to a remote JobNimbus MCP server using a lightweight, stateless MCP client. This setup lets you run a suite of remote tools against JobNimbus data without storing API keys on the server, while providing automatic rate limiting, health monitoring, and easy extensibility.

How to use

Install and run the MCP server locally or deploy it to a hosting provider. From Claude Desktop, configure the MCP server as a stdio connection by pointing to the local MCP client script. Use your JobNimbus API key and the target instance to authenticate each request. The server will act as a temporary proxy, forwarding requests to JobNimbus and returning results without persisting sensitive keys.

How to install

Prerequisites: ensure you have Node.js and npm installed on your machine. You may also need a hosting option if you plan to deploy remotely.

# 1. Clone the project
git clone https://example.com/benitocabrerar/jobnimbus-mcp-remote.git
cd jobnimbus-mcp-remote

# 2. Install dependencies
npm install

# 3. Copy environment configuration for local use
cp .env.example .env

# 4. Run the server in development mode
npm run dev

Open the health endpoint to verify the server is running.

http://localhost:3000/health

Additional configuration and usage notes

If you plan to connect Claude Desktop, you can configure a local stdio MCP connection using the example JSON shown below. This configuration uses a local node script as the MCP client and provides the necessary environment variables for authentication and target instance.

{
  "mcpServers": {
    "jobnimbus_stamford": {
      "type": "stdio",
      "command": "node",
      "args": ["C:/ruta/a/examples/mcp-client.js"],
      "env": {
        "MCP_SERVER_URL": "https://tu-servidor.onrender.com",
        "JOBNIMBUS_API_KEY": "tu_api_key_stamford",
        "JOBNIMBUS_INSTANCE": "stamford"
      }
    }
  }
}

Security and best practices

The MCP server extracts the API key from the request header, validates its format, uses it to call JobNimbus, and immediately clears the key from memory. It does not store API keys in any database or file, does not log them, and does not cache them.

Troubleshooting

If you cannot reach the health endpoint, verify that the server process is running, check network accessibility, and ensure the MCP URL configured in Claude Desktop matches the running server endpoint.

Available tools

validate_api_key

Validate the format and presence of the API key in incoming requests to ensure only well-formed keys are processed.

get_jobs

Fetch a list of jobs with optional pagination and filtering.

search_jobs

Search for jobs based on criteria such as keywords, status, or dates.

get_contacts

Retrieve contact records associated with jobs or leads.

get_invoices

Retrieve invoices related to jobs or clients.

get_attachments

Access attachments linked to specific jobs for review.

get_estimates

Fetch estimate data associated with a job or client.

get_calendar_activities

Obtain calendar events related to jobs, tasks, and appointments.