handsontable
Model Context Protocol server providing access to Handsontable documentation for AI-powered code generation
claude mcp add --transport stdio greenflux-handsontable-mcp-server node /absolute/path/to/handsontable-mcp-server/index.js
How to use
Handsontable MCP Server exposes three MCP tools to help AI agents access Handsontable documentation programmatically. The get_doc tool fetches specific documentation pages and returns them as Markdown, the search_docs tool lets you look up topics by keyword across all supported pages, and list_categories returns an overview of all documentation categories and topics. The server supports multiple frameworks (vanilla JavaScript, React, Angular, Vue 2/3) and includes safeguards such as input validation, rate limiting (one request every 100 ms), an LRU cache with a 1-hour TTL and a maximum of 100 items, and HTML-to-Markdown conversion for clean integration with downstream tools. To use the server, connect your MCP client (e.g., Claude Desktop) to the configured endpoint, then invoke get_doc, search_docs, or list_categories to retrieve Handsontable documentation data in a structured format suitable for AI-assisted code generation.
Typical workflows include requesting a specific documentation page (topic) in a chosen framework and type (guide or API) via get_doc, performing keyword searches with search_docs, or listing available categories with list_categories. The server fetches Handsontable HTML, parses and sanitizes it, converts it to Markdown, caches results, and returns the requested content in a clean, consistent structure for downstream consumption by your tooling chain.
How to install
Prerequisites:
- Node.js and npm installed on your system
- Basic shell/terminal familiarity
Installation steps:
- Clone the repository: git clone https://github.com/GreenFlux/handsontable-mcp-server.git
- Navigate to the project directory: cd handsontable-mcp-server
- Install dependencies: npm install
- Run the development server (optional for local testing): npm run dev
- If you plan to integrate with Claude Desktop, ensure the MCP server is reachable by your Claude config and reference the local path to the index.js file as shown in the Quick Setup. Example for Claude Desktop configuration is provided in the README under "Add to Claude Desktop".
Notes:
- You can also run the server directly with Node if you prefer not to use npm run dev, e.g., node index.js, depending on how the project is started in your environment.
- The README mentions a prebuilt Claude integration path and environment setup; adapt the command path to the actual installation path on your machine.
Additional notes
Tips and considerations:
- The MCP server enforces a 100 ms gap between requests to avoid overloading the upstream Handsontable site.
- It uses an LRU cache with a 1-hour TTL and a maximum of 100 cached items to speed up repeated requests.
- Results are converted from HTML to Markdown, with navigation and footer elements stripped to keep output clean for AI processing.
- Ensure Claude Desktop or any MCP client can reach the server endpoint; if you run locally, consider using a stable localhost port or a reachable private host in your environment.
- If you encounter rate-limit related errors, verify that your client is not issuing requests more rapidly than the configured 100 ms interval.
- The server supports multiple frameworks; specify the framework when using get_doc to fetch framework-specific documentation (e.g., javascript, react, angular).
- If you need to adjust behavior (cache size, TTL, rate limits), look for configuration options in the codebase or environment-specific overrides as documented in DEVELOPMENT notes.
Related MCP Servers
agent-toolkit
Collection of resources to help AI agents build better with Sanity.
mcp-rest-api
A TypeScript-based MCP server that enables testing of REST APIs through Cline. This tool allows you to test and interact with any REST API endpoints directly from your development environment.
claude-code
MCP Server connects with claude code local command.
MCPretentious
A powerful Model Context Protocol (MCP) server that puppeteers iTerm2 terminals.
Hybrid-CLI-Agent
Multi-agent CLI orchestrator combining Claude Code, Gemini CLI, and OpenRouter.
claude-image-gen
AI-powered image generation using Google Gemini, integrated with Claude Code via Skills or Claude.ai via MCP (Model Context Protocol).