mcp

This MCP server exposes a set of kintone-specific MCP tools that integrate with your AI workflows. The server itself runs the kintone MCP package, enabling tools like kintone-get-apps, kintone-get-app, and kintone-get-form-fields to be invoked in an MCP context. Use these tools to fetch app metadata, retrieve specific form field definitions, and assemble data for downstream AI processing. The server is designed to be run via an npm/npx workflow, and it consumes KINTONE_BASE_URL, KINTONE_USERNAME, and KINTONE_PASSWORD for authentication. You can run the server in a containerized or local environment and reference the tools in your MCP configuration or AI prompts.

Prerequisites:

• Node.js and npm (or npx available via npm install)
• Access to a kintone environment with a valid base URL, username, and password

Installation steps:

1. Install the MCP server package globally using npm:

npm install -g @kintone/mcp-server

2. Verify installation by running the MCP server tool (you can also use npx without global install):

npx -y @kintone/mcp-server --help

3. Run the MCP server in your environment by providing credentials (either via environment variables or command-line arguments). Example with environment variables:

export KINTONE_BASE_URL=https://example.cybozu.com
export KINTONE_USERNAME=username
export KINTONE_PASSWORD=password

kintone-mcp-server

4. If you prefer using npx directly for a one-off run, you can execute:

KINTONE_BASE_URL=https://example.cybozu.com KINTONE_USERNAME=username KINTONE_PASSWORD=password npx -y @kintone/mcp-server

Prerequisites explained:

• Node.js and npm are required to install and run the MCP server package.
• Ensure your kintone credentials have the necessary permissions for the operations you intend to perform with the tools.

Tips and notes:

• You can supply credentials via environment variables or pass them as command-line arguments to the server depending on your workflow.
• The MCP tools exposed by this server include: kintone-get-apps (list apps), kintone-get-app (details for a single app), and kintone-get-form-fields (form field metadata). Use these tools to prepare data for prompts and automations.
• If you operate behind a proxy, set HTTPS_PROXY in your environment to enable outbound HTTP requests.
• For security, avoid hard-coding credentials in scripts. Prefer environment variables or secret management where possible.
• When integrating with AI tooling, you can reference the MCP tools in prompts and feed the JSON outputs to downstream processors.