mcp-web-calc
LM Studio MCP server: web search (DuckDuckGo HTML + Playwright), Wikipedia, URL extract, and a precise calculator no API keys.
claude mcp add --transport stdio khanhs-234-mcp-web-calc npm run dev \ --env MAX_BYTES="20971520" \ --env USER_AGENT="mcp-web-calc/0.3 (https://local)" \ --env MAX_RESULTS="10" \ --env HTTP_TIMEOUT="15000" \ --env LANG_DEFAULT="vi"
How to use
mcp-web-calc is an MCP server for LM Studio that provides six built-in tools without requiring any API keys. It exposes search_web for two-tier web search (fast via DuckDuckGo HTML and deep via Playwright/Bing), fetch_url to load and extract content from URLs, HTML, or PDFs, summarize_url to fetch and briefly summarize content from a URL, math_eval for precise mathematical evaluation (including numbers, BigNumber, and fractions), wiki_get to retrieve a concise summary of a single Wikipedia page in a chosen language, and wiki_multi to fetch summaries across multiple languages in one call. These tools operate over MCP/stdio, enabling you to compose tasks and chain results within LM Studio without external authentication.
To use these tools, call each tool by name with the required input structures defined in the tool schema, or rely on LM Studio’s quick tests to evaluate small examples. For example, search_web can fetch a handful of relevant results with optional language and limit controls; fetch_url can extract text and metadata from a page or a PDF; and math_eval supports derivative, simplify, and other mathjs-based operations. The wiki tools provide lightweight access to summaries from Wikipedia in multiple languages, ideal for multilingual research tasks. The server is designed to run locally or within LM Studio, using the same MCP protocol as other servers, and does not require an API key or external service bindings.
Notes: ensure Playwright Chromium is installed for deep search mode. The server expects a Node.js 18+ environment and will bind through stdio with LM Studio, so run it via the provided command and configure the working directory accordingly.
How to install
Prerequisites:
- Node.js 18+ installed on your system
- Supported OS: Windows, macOS, Linux
- ít required: Playwright browser binaries for deep search (Chromium)
Installation steps:
# 1) Install dependencies
npm i
# 2) Install Playwright Chromium (needed for deep search mode)
npx playwright install chromium # cần cho 'deep' (Bing/Playwright)
# 3) Run in development mode
npm run dev
# Optional: Build for production and start
npm run build && npm run start
Environment setup (recommended):
# In LM Studio or a .env file in your project root
USER_AGENT=mcp-web-calc/0.3 (https://local)
HTTP_TIMEOUT=15000 # ms cho mọi yêu cầu mạng
MAX_RESULTS=10 # mặc định cho search_web
LANG_DEFAULT=vi # ngôn ngữ mặc định
MAX_BYTES=20971520 # giới hạn tải 20MB cho fetch_url
Khi chạy thành công, log sẽ hiển thị: mcp-web-calc ready (stdio)…
Additional notes
Tips and troubleshooting:
- CAPTCHA or temporary blocks: reduce request rate and retry later. Prefer mode: fast when needed.
- If Playwright is not installed: run npx playwright install chromium.
- Timeouts or long fetches: increase HTTP_TIMEOUT or reduce MAX_BYTES if content is large.
- Internal URLs blocked safeguards (SSRF guard): the server blocks localhost/127.0.0.1/::1, etc., by default.
- In LM Studio, add the server under Settings → Developer → Model Context Protocol (MCP) Servers and provide the same environment variables as above. The log line to confirm readiness is: mcp-web-calc ready (stdio).
Related MCP Servers
context7
Context7 MCP Server -- Up-to-date code documentation for LLMs and AI code editors
obsidian -tools
Add Obsidian integrations like semantic search and custom Templater prompts to Claude or any MCP client.
MiniMax -JS
Official MiniMax Model Context Protocol (MCP) JavaScript implementation that provides seamless integration with MiniMax's powerful AI capabilities including image generation, video generation, text-to-speech, and voice cloning APIs.
mcp-bundler
Is the MCP configuration too complicated? You can easily share your own simplified setup!
akyn-sdk
Turn any data source into an MCP server in 5 minutes. Build AI-agents-ready knowledge bases.
promptboard
The Shared Whiteboard for Your AI Agents via MCP. Paste screenshots, mark them up, and share with AI.