Lark Dashboard

You can run an MCP server that lets you harness Claude Code to automate creating and managing Lark dashboards from your TypeScript workflow. This MCP setup enables you to run a local server, expose your dashboard-building capabilities, and compose complex dashboards through a defined set of tools and blocks.

How to use

Install the package and wire up your MCP client to your Lark dashboard workflow. Use the MCP server to expose commands that create dashboards, add blocks (charts, metrics, views, text, lists, and more), and manage existing dashboards. You can also run the local MCP server with environment variables that supply your API keys and region so Claude Code can authenticate and operate on your behalf.

How to install

Prerequisites: Node.js (version 16 or newer) and npm, or Python 3.8+ if you use the Python SDK in parallel workflows.

Install the TypeScript SDK package.

Set up the MCP server configuration locally to enable Claude Code integration.

MCP Server Usage

The MCP server is configured to run as a stdio server using the following command. This enables Claude Code to interact with your hype-dash features directly from your development environment.

{
  "mcpServers": {
    "hype_dash": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@hypelab/hype-dash"],
      "env": {
        "LARK_API_KEY": "your-api-key-here",
        "LARK_REGION": "sg"
      }
    }
  },
  "env": [
    {"name": "LARK_API_KEY", "description": "API key for Lark", "required": true, "example": "abc123"},
    {"name": "LARK_REGION", "description": "Lark region (sg/cn/us)", "required": true, "example": "sg"}
  ]
}

Available MCP Tools

The MCP server exposes a set of tools to manage dashboards and blocks programmatically. Use these tools to build and modify dashboards without leaving your code editor.

Configuration

Client options in your TypeScript project let you customize how you connect to the Lark API and how you handle logging and retries. You can set your API key, region, base API URL, and various timeouts and retry policies.

Environment variables required for MCP operation include your Lark API key and the target region. These are used by the MCP server to authenticate requests and route them to the correct workspace.

Advanced Features

Batch operations let you create multiple blocks in a single call for efficiency. Filtering and error handling help you compose complex dashboards while gracefully handling validation or API errors.

Troubleshooting

Authentication errors usually mean the API key is incorrect or lacks necessary permissions. Network issues often point to firewall rules or wrong region configuration. Validation errors indicate missing required fields or mismatched data types.

Examples

You can build a dashboard with a bar chart, a metrics card, and a tab view by chaining builder calls in your TypeScript code and then sending the blocks to the MCP server for creation.

Available tools

create_dashboard

Create a new dashboard with a given name and app token.

create_chart_block

Add a chart block to a dashboard with a specified data source and configuration.

create_metrics_block

Add a metrics block to display KPIs like sums, averages, or counts.

create_view_block

Add a data view such as a table or Kanban view to a dashboard.

create_text_block

Add rich text blocks to provide context or explanations on a dashboard.

list_dashboards

List all dashboards available to the configured API key.

delete_dashboard

Remove a dashboard by its ID.