mcp-web-calc
LM Studio MCP server: web search (DuckDuckGo HTML + Playwright), Wikipedia, URL extract, and a precise calculator no API keys.
Install / Use
claude mcp add khanhs-234 -- npx -y github:khanhs-234/mcp-web-calcIf the server publishes to npm under a different name, use that package instead — check the repo README.
MCP Server
Model Context Protocol server
Quality Score
Category
Development & EngineeringSupported Platforms
Our assessment of mcp-web-calc
mcp-web-calc scores 68/100 on our quality scale, 3852nd of 4,614 Development & Engineering skills we index.
Its MCP Server is 5.2 KB long, well organised into 17 sections with 19 code examples: a solid amount of guidance for an agent.
It has 3 GitHub stars, so there is little community track record yet; judge it on its content.
Maintenance, license and trust
- The repository was last updated about 13 months ago. Expect some instructions to reference tool versions or APIs that have since changed.
- Our last check on 2026-10-19 found the source still online.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 80/100, with 2 cautions from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-09-24. Automated pattern scan on 2026-09-24. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
mcp-web-calc compared with similar skills
All 4 of these similar skills score higher than mcp-web-calc; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| mcp-web-calc (this skill)by khanhs-234 | 68 | 3 | 13mo ago | MCP Server |
| taste-skillby Leonxlnx | 100 | 111.0k | 1d ago | CLAUDE.md |
| Agent-Reachby Panniantong | 100 | 109.5k | 1d ago | CLAUDE.md |
| headroomby headroomlabs-ai | 100 | 79.0k | today | CLAUDE.md |
| ai-job-searchby ai-job-search-dev | 100 | 70.0k | today | CLAUDE.md |
Frequently asked questions
- How do I install mcp-web-calc?
- Run
claude mcp add khanhs-234 -- npx -y github:khanhs-234/mcp-web-calc. The install tabs above show the steps for each supported agent. - Which AI agents does mcp-web-calc work with?
- It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
- Is mcp-web-calc safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 80/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is mcp-web-calc still maintained?
- The repository was last updated about 13 months ago. Expect some instructions to reference tool versions or APIs that have since changed.
Skill content
View source on GitHubmcp-web-calc — MCP server cho LM Studio (không cần API key)
Phiên bản: 0.3.0
6 công cụ tích hợp
search_web— Tìm web hai tầng (Nhanh: DuckDuckGo HTML → Sâu: Playwright/Bing)fetch_url— Tải & trích xuất nội dung từ URL/HTML/PDF (Readability + pdf-parse)summarize_url— Lấy nội dung từ URL rồi tóm tắt ngắn gọnmath_eval— Máy tính chính xác (Number / BigNumber / Fraction)wiki_get— Lấy tóm tắt 1 trang Wikipedia theo ngôn ngữwiki_multi— Lấy tóm tắt Wikipedia theo nhiều ngôn ngữ trong một lần gọi
Không dùng API key. Hoạt động hoàn toàn qua MCP/stdio.
Yêu cầu
- Node.js 18+
- Windows/macOS/Linux
- (Tìm kiếm sâu) Playwright Chromium
Cài đặt
npm i
npx playwright install chromium # cần cho 'deep' (Bing/Playwright)
# Dev:
npm run dev
# hoặc build & chạy prod:
npm run build && npm run start
Biến môi trường (khuyến nghị)
Đặt trong LM Studio hoặc .env:
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
Bảo vệ SSRF:
fetch_urlsẽ chặnlocalhost/127.0.0.1/::1/.local/.localhost.
Thêm MCP server trong LM Studio
Settings → Developer → Model Context Protocol (MCP) Servers → Add
- Name:
mcp-web-calc - Command:
npm - Args:
run,dev(hoặcstartnếu đã build) - Working directory: đường dẫn tới thư mục dự án
- Environment variables: như ở phần trên
Khi chạy thành công, log sẽ hiện: mcp-web-calc ready (stdio)…
Công cụ & cú pháp (tool schema)
1) search_web
- Input
{
q: string;
limit?: number; // default: MAX_RESULTS (env), clamp 1..50
lang?: string; // "vi" | "en" | ...
mode?: "fast"|"deep"|"auto"; // default: "auto"
}
- Output (rút gọn)
{
items: { title: string; url: string; snippet?: string; source: "ddg_html"|"bing_pw" }[];
modeUsed: "fast"|"deep"|"auto";
enginesUsed: ("ddg_html"|"bing_pw")[];
escalated: boolean; // auto có leo sang deep hay không
diagnostics?: Record<string, unknown>;
}
- Ví dụ
search_web: { "q": "Node.js LTS release schedule", "mode": "fast", "limit": 5, "lang": "vi" }
2) fetch_url
- Input
{ url: string } // hỗ trợ HTML & PDF
- Output (rút gọn)
{
text: string;
url: string;
title?: string;
byline?: string;
siteName?: string;
lang?: string;
length?: number; // nếu có
}
- Ví dụ
fetch_url: { "url": "https://example.com" }
3) summarize_url
- Input
{ url: string }
- Hành vi
- Lấy nội dung bằng
fetch_url, sau đó thử gọi model phía server (nếu được LM Studio cấp) để tóm tắt ngắn gọn tiếng Việt. - Fallback: nếu không gọi được model, trả về tối đa ~2000 ký tự đầu của bài.
- Lấy nội dung bằng
- Ví dụ
summarize_url: { "url": "https://developer.mozilla.org/en-US/docs/Web/API/Fetch_API" }
4) math_eval
- Input
{
expression: string; // cú pháp mathjs
mode?: "number"|"BigNumber"|"Fraction"; // default: "BigNumber"
precision?: number; // default: 64 (khi BigNumber)
}
- Lưu ý
mathjskhông có hàmintegrate. Dùngderivative,simplify, ma trận, đơn vị, v.v.- Cần lượng giác? Dùng
mode="number"để có sin/cos… chính xác kiểu số JS.
- Ví dụ
math_eval: { "expression": "derivative(sin(x)*exp(x), x)", "mode": "number" }
5) wiki_get
- Input
{ title: string; lang?: string /* default "vi" */ }
- Output (rút gọn)
{ lang: string; title: string; url: string; description?: string; extract?: string; thumbnailUrl?: string }
- Ví dụ
wiki_get: { "title": "Việt Nam", "lang": "vi" }
6) wiki_multi
- Input
{
term: string, // thuật ngữ gốc, ví dụ "Cá"
baseLang?: string, // mặc định "vi"
langs?: string[] // mặc định ["vi","en"]
}
- Output (rút gọn)
{
baseLang: string;
base: { /* WikiSummary của baseLang */ };
items: Record<string, WikiSummary | null>; // null nếu không tìm thấy
resolved: Record<string, { title?: string; source: "base"|"langlinks"|"direct"|"none" }>
}
- Ví dụ
wiki_multi: { "term": "Cá", "baseLang": "vi", "langs": ["vi","en","ja","fr"] }
Mẹo kiểm thử nhanh
search_web: { "q": "site:developer.apple.com App Intents", "mode": "deep", "limit": 5, "lang": "vi" }
fetch_url: { "url": "https://example.com" }
summarize_url: { "url": "https://www.python.org/dev/peps/pep-0008/" }
math_eval: { "expression": "inv([[1,2],[3,4]])", "mode": "number" }
wiki_get: { "title": "Lambda calculus", "lang": "en" }
wiki_multi: { "term": "Cá", "baseLang": "vi", "langs": ["vi","en","ja","fr"] }
Khắc phục sự cố
- CAPTCHA / chặn tạm thời: giảm tần suất, ưu tiên
mode="fast", thử lại sau vài phút. - Playwright chưa cài:
npx playwright install chromium. - Timeout / treo khi tải: tăng
HTTP_TIMEOUT, hoặc nội dung vượtMAX_BYTES. - URL nội bộ bị chặn: đây là chủ đích (SSRF guard).
Giấy phép
MIT — xem tệp LICENSE.
Related Skills
taste-skill
111.0kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
Agent-Reach
109.5kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
headroom
79.0kCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
ai-job-search
70.0kThe job search that runs on your machine. AI job application framework built on Claude Code: evaluate postings, tailor CVs, write cover letters, prep interviews. Fork it and own it.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
