{"id":"a710a626-b039-4365-91e4-0ed7f91d2628","slug":"smithery-devshero-shadowcrawl","name":"🥷 ShadowCrawl: The Zero-Docker \"Unstoppable\" Stealth Scraper & Search","description":"**Pure Rust MCP Server**\n\nShadowCrawl is a high-performance, Zero-Docker MCP server written in Rust. It serves as a 100% private, sovereign alternative to Firecrawl, Jina Reader, and Tavily.\n\nUnlike other scrapers, ShadowCrawl v2.3.0 runs as a single standalone binary with native Chromium control (CDP) and a Human-In-The-Loop (HITL) fallback system, ensuring you can bypass 99.9% of bot protections (Cloudflare, DataDome) without complex infrastructure.\n\nWhy AI Agents love ShadowCrawl:\n\n- Zero-Docker & Zero-Config: No Redis, No Qdrant, No SearXNG. Single binary setup.\n- God-Tier Anti-Bot Bypass: Native Chromiumoxide (CDP) with JS-level stealth injection and HITL (Human-In-The-Loop) fallback for solving CAPTCHAs.\n- Internal Meta-Search: Parallel search across Google, Bing, DuckDuckGo, and Brave without external API keys.\n- Smart Ad-Blocking: Built-in high-speed aho-corasick engine to strip ads and trackers before extraction.\n- Semantic Memory: Embedded LanceDB for 100% local, private research history recall.\n- AI-Optimized Markdown: Delivers ultra-clean content stripped of \"Buddhist Era\" dates and web noise.\n\n\nTools list:\n\nsearch_web — federated search (No API Key needed)\nsearch_structured — search + top result scraping\nscrape_url — single URL extraction\nscrape_batch — multi-URL parallel scraping\ncrawl_website — bounded recursive crawling\nextract_structured — schema-driven extraction\nresearch_history — semantic recall from prior runs\nproxy_manager — proxy list/status/switch/test/grab operations\nnon_robot_search — [NEW] The \"Nuclear Option\" for Boss-level anti-bots (LinkedIn/Cloudflare) with HITL.","capabilities":[],"protocols":["MCP"],"safetyScore":86,"overallRank":33.9,"trustScore":null,"trust":null,"source":"SMITHERY","updatedAt":"2026-04-15T00:48:45.152Z"}