home / mcp / ai connect mcp server

AI Connect MCP Server

An MCP (Model Context Protocol) server that allows AI agents to query and manage jobs in the Agent Jobs system of the AI Connect platform.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "aiconnect-cloud-agentjobs-mcp": {
      "command": "npx",
      "args": [
        "@aiconnect/agentjobs-mcp"
      ],
      "env": {
        "DEFAULT_ORG_ID": "your-organization",
        "AICONNECT_API_KEY": "your-api-key",
        "AICONNECT_API_URL": "https://api.aiconnect.cloud/api/v0"
      }
    }
  }
}

You run the AI Connect MCP Server to expose a standardized interface for AI agents to manage AI Connect Jobs. With this server, agents can list, query, create, cancel, and monitor jobs, enabling asynchronous workflows across channels and systems.

How to use

Use an MCP client to perform common job operations against the AI Connect MCP Server. You can list all jobs with filters, retrieve a single job by ID, create new jobs for immediate or scheduled execution, cancel running or scheduled jobs, and monitor overall status (WAITING, RUNNING, COMPLETED, FAILED, CANCELED). Typical workflows include starting a scheduled job, checking its progress, or canceling a job that is no longer needed.

How to install

Prerequisites: you need Node.js and npm installed on your machine. You can verify by running node -v and npm -v in your terminal.

Step 1: Run the MCP server directly with NPX (recommended) to try it out without installation.

npx @aiconnect/agentjobs-mcp --help

Step 2: For a local, persistent setup, clone the project, install dependencies, and configure environment variables.

git clone <repository-url>
cd agentjobs-mcp
npm install

Step 3: Configure environment variables. Copy the example and edit values as needed. If you omit values, sensible defaults apply, but you must provide an API key for API calls to work.

cp .env.example .env
```
Edit .env with your credentials, for example:

```
DEFAULT_ORG_ID=your-organization
AICONNECT_API_KEY=your-api-key
AICONNECT_API_URL=https://api.aiconnect.cloud/api/v0

Step 4: Build the project before running in a development or production environment.

npm run build

Step 5: Start the MCP server locally.

npm start

Configuration and start tips

The server supports configuration via environment variables or defaults. If AICONNECT_API_KEY or AICONNECT_API_URL are not provided, you will receive clear guidance on what to configure. Default organization ID is aiconnect if not specified.

To use Decimal CLI or automate startup in a script, you can rely on the NPX invocation for ad-hoc runs or the npm start script for a more stable, server-style process.

Security and credentials

Protect your API key and restrict access to agents that should manage AI Connect Jobs. Store credentials securely (for example, in a protected environment or secret manager) and avoid embedding keys in client-side code.

Examples of common operations

You can create a new job for immediate or scheduled execution, list jobs with filters, retrieve a single job, or cancel a job if needed. These actions enable automated workflows and cross-channel job management.

Available tools

get_jobs_stats

Fetch aggregated statistics for agent jobs, useful for dashboards and monitoring with filtering by time, org, job type, tags, status, and channel.

list_jobs

List all jobs with filtering and pagination options including status, job type, channel, limit, offset, and sorting.

get_job

Retrieve details of a specific job by its ID.

create_job

Create a new job for immediate or scheduled execution, with configuration for timeouts, retries, and parameters.

cancel_job

Cancel a running or scheduled job, with an optional reason.