MCP Servers

A collection of Model Context Protocol servers, templates, tools and more.

K
Kinaxis MCP Server

MCP server by federicociro

Created 6/16/2026
Updated about 17 hours ago
Repository documentation and setup instructions

kinaxis-mcp-server

License: MIT

The first open-source Model Context Protocol (MCP) server for Kinaxis Maestro (formerly RapidResponse).

What it does

Exposes Kinaxis Maestro REST & Bulk APIs as MCP tools that AI agents (Copilot, Claude, etc.) can call to:

| Tool | Description | |------|-------------| | health_check | Verify connectivity and auth | | list_tables | Discover available Kinaxis tables | | query_table | Run a query against any table | | fetch_query_results | Paginated row retrieval | | close_query | Release query resources | | bulk_export | Start a bulk export from any table | | fetch_export_results | Paginated row retrieval from bulk export | | bulk_upload_init | Start a bulk data upload | | bulk_upload_data | Push rows to an upload | | bulk_upload_complete | Finalize upload & trigger update | | trigger_data_update | Trigger a data update | | data_update_status | Check data update progress | | run_script | Execute Maestro scripts (sync) | | run_script_async | Execute scripts asynchronously | | run_workflow | Run Maestro workflows | | workflow_status | Check workflow status | | open_workbook | Open a workbook for data retrieval | | read_worksheet | Read worksheet data | | import_worksheet | Import data via workbook |

Quick start

Prerequisites

  • Node.js >= 18
  • A Kinaxis Maestro instance with REST API access
  • OAuth2 client credentials or a web service user (Basic Auth)

Install

git clone https://github.com/federicociro/kinaxis-mcp-server.git
cd kinaxis-mcp-server
npm install
cp .env.example .env
# Edit .env with your Kinaxis credentials

Configure

Edit .env:

KINAXIS_BASE_URL=https://your-region.kinaxis.net/YOUR_INSTANCE
KINAXIS_CLIENT_ID=your_oauth2_client_id
KINAXIS_CLIENT_SECRET=your_oauth2_client_secret
KINAXIS_DEFAULT_SCENARIO=Baseline
KINAXIS_DEFAULT_SCOPE=Public

Or use Basic Auth:

KINAXIS_BASE_URL=https://your-region.kinaxis.net/YOUR_INSTANCE
KINAXIS_USERNAME=your_ws_user
KINAXIS_PASSWORD=your_ws_password

Add to your MCP client

VS Code (Copilot)

Add to .vscode/mcp.json:

{
  "servers": {
    "kinaxis": {
      "command": "node",
      "args": ["src/index.js"],
      "cwd": "/path/to/kinaxis-mcp-server",
      "env": {
        "KINAXIS_BASE_URL": "https://your-region.kinaxis.net/YOUR_INSTANCE",
        "KINAXIS_CLIENT_ID": "your_client_id",
        "KINAXIS_CLIENT_SECRET": "your_client_secret"
      }
    }
  }
}

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "kinaxis": {
      "command": "node",
      "args": ["/path/to/kinaxis-mcp-server/src/index.js"],
      "env": {
        "KINAXIS_BASE_URL": "https://your-region.kinaxis.net/YOUR_INSTANCE",
        "KINAXIS_CLIENT_ID": "your_client_id",
        "KINAXIS_CLIENT_SECRET": "your_client_secret"
      }
    }
  }
}

Run standalone

npm start

Authentication

The server supports two authentication methods:

  1. OAuth2 Client Credentials (recommended): Register a client in Kinaxis Admin, provide KINAXIS_CLIENT_ID and KINAXIS_CLIENT_SECRET. Tokens are refreshed automatically.

  2. Basic Auth: Use a Kinaxis web service user with KINAXIS_USERNAME and KINAXIS_PASSWORD. Simpler setup but credentials are sent with every request.

Kinaxis API reference

This server implements endpoints from the Maestro Web Services Guide (REST & Bulk APIs), available on the Kinaxis Customer Community. Place your copy of the documentation in the docs/ folder (gitignored) for local reference.

Contributing

PRs welcome! Please:

  1. Fork the repo
  2. Create a feature branch (feat/my-feature)
  3. Use Conventional Commits
  4. Open a PR

See CONTRIBUTING.md for local setup, commit examples, and release notes guidance.

License

MIT Federico Berrone, 2026

Quick Setup
Installation guide for this server

Install Package (if required)

npx @modelcontextprotocol/server-kinaxis-mcp-server

Cursor configuration (mcp.json)

{ "mcpServers": { "federicociro-kinaxis-mcp-server": { "command": "npx", "args": [ "federicociro-kinaxis-mcp-server" ] } } }