MCP Web Research Server
Provides a Google search tool for executing organic web searches through stealth Chrome, including consent-dialog bypass, canonicalized result URLs, rich snippet extraction, and filters for result limit, language, country, and SafeSearch.
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
π MCP Web Research Server
Enterprise-Grade, Ultra-Fast, and Production-Ready Web Research Model Context Protocol (MCP) Server.
Engineered for AI coding agents and autonomous workflows requiring high-accuracy Google Search and high-density, token-efficient webpage content extraction with undetectable anti-bot evasion.
π Key Architectural Features
π Undetectable Anti-Bot Evasion (Patchright + Chrome): Built on Patchright (patched driver eliminating
Runtime.enable,--enable-automation, and console leaks). Enforces official Google Chrome (channel: 'chrome') with zero-fallback for 100% genuine consumer browser fingerprints.β‘ Pre-Warmed Persistent Pool (<100ms Tab Acquisition): Browser context is pre-warmed at server boot with a persistent user profile (
launchPersistentContext), retaining DNS, TLS session tickets, V8 bytecode, and consent cookies for 38β50% faster navigations.π Smart Progressive Auto-Scroll & DOM Convergence: Handles complex e-commerce, SPA, and XHR-heavy sites (e.g. Tokopedia, Shopee, Reddit). Progressively scrolls, triggers
IntersectionObserverlazy-loads, auto-clicks content expanders ("Lihat Selengkapnya" / "Read more"), and stops when DOM height converges.π Token-Efficient & High-Density Markdown: Strips token-wasting avatar images, promotional widgets, and redundant blank lines. Compiles clean GitHub Flavored Markdown via Unified AST (
linkedom+@mozilla/readability+rehype-remark+remark-gfm).π‘οΈ Zero-Trust SSRF Defense: Rejects loopback (
127.0.0.0/8,::1), RFC 1918 private subnets (10.0.0.0/8,172.16.0.0/12,192.168.0.0/16), link-local metadata (169.254.169.254,metadata.google.internal), and non-HTTP protocols.π₯οΈ Cross-Platform Isolation: Dedicated virtual framebuffer (
DISPLAY=:99) on Linux; automatic background stealth headless mode on macOS and Windows to prevent desktop window focus-stealing.π₯ Real Liveness & Readiness Probing: Provides real subsystem health metrics over both MCP STDIO (
healthCheck) and native HTTP REST (GET /healthwith HTTP 200/503 status codes for Kubernetes / Docker).
Related MCP server: Crawlberg MCP Server
π₯οΈ Operating System Compatibility Matrix
Operating System | Support Level | Windowing & Display | GPU / WebGL Backend | Notes |
Linux (Ubuntu 24.04 / Debian) | β TESTED & VERIFIED | Xvfb Virtual Framebuffer ( | SwiftShader CPU ( | Fully tested in production CI and local environments. Zero physical screen pop-ups. |
macOS ( | β οΈ UNTESTED (Architecturally Supported) | Quartz / WindowServer | Native Apple Metal | Headless defaults to |
Windows ( | β οΈ UNTESTED (Architecturally Supported) | Desktop Window Manager (DWM) | Native Direct3D11 | Headless defaults to |
Notice: Linux (Ubuntu 24.04 LTS) is the primary verified operating environment. macOS and Windows contain platform-adaptive code paths, but are currently classified as untested / experimental.
π Prerequisites & System Requirements
Node.js: v20.x or v22.x+ (tested on Node v24.16.0)
Package Manager:
pnpm(v9+ or v12+)Google Chrome: Official Google Chrome binary installed on host:
npx patchright install chrome # Or install Google Chrome via standard apt/rpm packageXvfb (Linux only): Virtual display running on
:99(recommended systemd unit below).
Example Linux Xvfb Systemd Service (/etc/systemd/system/xvfb.service)
[Unit]
Description=Single-Instance X Virtual Framebuffer (Display :99)
After=network.target systemd-tmpfiles-setup.service
Requires=systemd-tmpfiles-setup.service
[Service]
Type=simple
User=xvfb
Group=xvfb
ExecStart=/usr/bin/Xvfb :99 -screen 0 1920x1080x24 -nolisten tcp -ac +extension GLX +render -noreset
ExecStartPost=/bin/sh -c 'for i in $(seq 1 30); do /usr/bin/xdpyinfo -display :99 >/dev/null 2>&1 && exit 0; sleep 0.2; done; exit 1'
Restart=always
RestartSec=2
StandardOutput=null
StandardError=journal
[Install]
WantedBy=multi-user.targetβοΈ 12-Factor Configuration Reference
All settings can be configured via environment variables. Universal defaults are applied automatically if unset:
Variable | Type | Default | Description |
|
|
| X11 virtual display target for Linux automation. |
|
|
| Browser headless flag. Defaults to |
|
|
| Optional HTTP/HTTPS proxy URI (e.g. |
|
|
| Persistent directory storing Chrome cache, cookies, and session state. |
|
|
| Default navigation timeout in milliseconds. |
|
|
| Maximum simultaneous browser tabs allowed by semaphore. |
|
|
| Lifetime requests before background browser restart to prevent RAM fragmentation. |
|
|
| Upper bound memory guard against massive HTML payloads. |
|
|
| Maximum wait time in ms for initial network stabilization before scrolling. |
|
|
| Debounce delay in ms after each scroll step (increase to 1000-2000 for slow networks). |
|
|
| Distance in pixels per scroll increment. |
|
|
| Final quiet period in ms after scroll loop completes. |
|
|
| Default maximum scroll loop iterations before convergence termination. |
|
|
| Maximum wait time in ms for Google SERP cards to render. |
|
|
| Minimum character length for Mozilla Readability article scoring. |
π Client Installation & Setup Guide
1. Claude Desktop
Add to your claude_desktop_config.json:
macOS:
~/Library/Application Support/Claude/claude_desktop_config.jsonWindows:
%APPDATA%\Claude\claude_desktop_config.jsonLinux:
~/.config/Claude/claude_desktop_config.json
{
"mcpServers": {
"web-research": {
"command": "node",
"args": ["/absolute/path/to/mcp-web-research/dist/entries/mcp-stdio.js"],
"env": {
"BROWSER_DISPLAY": ":99",
"SCROLL_DELAY_MS": "500",
"MAX_SCROLL_STEPS": "6"
}
}
}
}2. Cursor IDE
Add to ~/.cursor/mcp.json (or via Cursor Settings > Features > MCP):
{
"mcpServers": {
"web-research": {
"command": "node",
"args": ["/absolute/path/to/mcp-web-research/dist/entries/mcp-stdio.js"],
"env": {
"BROWSER_DISPLAY": ":99"
}
}
}
}3. OpenCode
Add to your opencode.json (or profile config):
{
"$schema": "https://opencode.ai/config.json",
"mcp": {
"web-research": {
"type": "local",
"command": ["node", "/absolute/path/to/mcp-web-research/dist/entries/mcp-stdio.js"],
"environment": {
"BROWSER_DISPLAY": ":99"
}
}
}
}4. VS Code (Cline / Roo Code / Claude Dev)
Add to your MCP Settings configuration file:
{
"mcpServers": {
"web-research": {
"command": "node",
"args": ["/absolute/path/to/mcp-web-research/dist/entries/mcp-stdio.js"],
"env": {
"BROWSER_DISPLAY": ":99"
},
"disabled": false,
"autoApprove": []
}
}
}5. Windsurf IDE
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"web-research": {
"command": "node",
"args": ["/absolute/path/to/mcp-web-research/dist/entries/mcp-stdio.js"],
"env": {
"BROWSER_DISPLAY": ":99"
}
}
}
}6. HTTP / Self-Hosted Deployment (Docker / Kubernetes)
Run as a standalone network service:
# Build the project
pnpm run build
# Start the Streamable HTTP server on port 3000
pnpm run startMCP Endpoint:
http://localhost:3000/mcp(Streamable HTTP transport)Kubernetes / Docker Healthcheck:
GET http://localhost:3000/health(Returns HTTP 200 / 503 with subsystem metrics)
π οΈ MCP Tools Catalog
1. google_search (googleSearchAction)
Executes organic web search on Google using stealth Chrome with automatic consent dialog bypass, breadcrumb canonicalization, and rich snippet extraction.
Input Parameters:
query(string, required, 1..500 chars): Search keywords or phrases.limit(integer, optional, 1..50, default: 10): Maximum organic results.language(string, optional, default: "en"): Language code (e.g."en","id").country(string, optional, default: "us"): Geolocation region code (e.g."us","id").safeSearch(boolean, optional, default: true): SafeSearch adult filter.
Example Tool Response:
{
"query": "typescript documentation",
"totalReturned": 2,
"results": [
{
"title": "The starting point for learning TypeScript",
"url": "https://www.typescriptlang.org/docs",
"redirectUrl": "https://www.google.com/goto?url=...",
"snippet": "TypeScript extends JavaScript by adding types to the language..."
}
],
"timestamp": "2026-11-20T00:00:00.000Z"
}2. web_extract (webExtractAction)
Navigates to any target URL, executes intelligent progressive scrolling to trigger lazy XHR/SPA content (prices, reviews, specs), auto-clicks expanders ("Lihat Selengkapnya"), and serializes dense GitHub Flavored Markdown.
Input Parameters:
url(string, required, format: "uri", 1..2048 chars): Target website URL (http://orhttps://).smartScroll(boolean, optional, default: true): Enable automated progressive scroll and XHR stabilization.scrollDelayMs(integer, optional, 100..5000, default: 500): Jeda debounce scroll per langkah. Naikkan ke 1000β2000ms jika koneksi internet lambat.maxScrolls(integer, optional, 0..20, default: 6): Batas iterasi scroll. Naikkan untuk halaman katalog atau ulasan yang sangat panjang.timeoutMs(integer, optional, 1000..60000, default: 30000): Maximum navigation deadline.
Example Tool Response:
{
"url": "https://www.tokopedia.com/kyo-official/ssd-kyo-kaizen-m-2-nvme-512gb-pcie-gen3-x4-2280-ssd-nvme-gen-3",
"title": "SSD KYO KAIZEN M.2 NVMe 512GB PCIe Gen3 x4 2280 SSD NVMe Gen 3",
"byline": "KYO Indonesia",
"excerpt": "KAIZEN diambil dari bahasa Jepang untuk perbaikan berkelanjutan...",
"markdown": "## SSD KYO KAIZEN M.2 NVMe 512GB...\n\n**Harga:** Rp1.626.050\n**Stok:** 451\n\n### Spesifikasi Lengkap:\n- Tipe: M.2 NVMe PCIe 3 x4\n- Kapasitas: 512GB\n\n## ULASAN PEMBELI\n- **Andri:** Barang diterima dengan baik dan cepat...",
"timestamp": "2026-11-20T00:00:00.000Z"
}3. healthCheck (healthAction)
Inspects real-time subsystem readiness, browser pool connectivity, active concurrency, and memory consumption.
Example Output:
{
"status": "healthy",
"timestamp": "2026-11-20T00:00:00.000Z",
"uptimeSec": 1240,
"browser": {
"connected": true,
"prewarmed": true,
"activeTabs": 0,
"maxCapacity": 5,
"requestCount": 42,
"userDataDir": "/home/shukoi/.cache/mcp-web-research/chrome-profile"
},
"display": ":99",
"memory": {
"heapUsedMb": 64.25,
"rssMb": 182.10
}
}π§ͺ Testing & Verification
Run the full automated verification pipeline:
# Execute unit and live integration tests (56+ tests)
pnpm test
# Verify strict TypeScript type-safety (0 errors)
pnpm run typecheck
# Check formatting and static quality (Biome)
pnpm run lint
# Compile production build to ./dist
pnpm run buildπ License
MIT Β© Shukoi Lab
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelogβ¦
This server cannot be deployed
Maintenance
Related MCP Connectors
Web search, Google SERP, page extraction to markdown and cited answers for AI agents.
Web search, browser automation, scraping, crawling and CAPTCHA solving for AI agents.
Web scraping, search, crawling, structured extraction and research for AI agents.
Web search and any page as LLM-ready Markdown for AI agents. Free keyless tier, no signup.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI agents to perform stealth web searches, browse pages with anti-bot bypass, and extract clean Markdown content, supporting direct image URLs and deep research.MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to crawl and scrape websites, converting HTML to clean Markdown and structured metadata with support for JavaScript rendering, bot evasion, and SSRF protection.197MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to perform live web searches with ranked results and extracted page passages, plus clean markdown extraction of any URL.MIT
- FlicenseNot gradedqualityCmaintenanceHigh-speed Google search intelligence and token-optimized Markdown grounding engine for AI LLMs, autonomous agents, and RAG pipelines.-