netlens-mcp

An MCP server for unobstructed web reading: it fetches any URL directly with
browser-like headers — past robots.txt and naive bot blocks — and returns the
full page as clean, ad-stripped Markdown, not a summary.
Requirements
- Python 3.10+ installed and on PATH
Quick Start
{
"mcpServers": {
"netlens": {
"command": "npx",
"args": ["-y", "netlens-mcp"]
}
}
}
What This Package Does
This npm package bundles the full NetLens server (pure Python, no pip install and
no network dependency after install). It:
- Checks that Python 3.10+ is installed
- Launches the bundled MCP server over stdio
Zero external Python dependencies — the server uses only the standard library.
Tools
web_search — search the web and get real result links (title/URL/snippet),
not summaries. Selectable engine (DuckDuckGo → Mojeek → Bing, or SearXNG).
web_fetch — fetch any page as full, clean Markdown. Modes: article
(main content, default), full (whole body), raw (HTML).
Configuration
NETLENS_SEARCH_ENGINE | auto | Default search backend: auto, duckduckgo, bing, mojeek, searxng |
NETLENS_SEARXNG_URL | — | Base URL of a SearXNG instance for engine=searxng |
Documentation
Full documentation and tool reference: GitHub