Wet Mcp

Gives Claude five search strategies, content extraction with automatic escalation from basic HTTP to headless browsing, and built-in library documentation indexing. The search tool hits SearXNG metasearch plus academic sources like Scholar and arXiv with cached results and citation formatting. Extract runs a multi-tier scraping chain that falls back to Crawl4AI when sites block simple requests, returning clean markdown plus structured JSON-LD. The docs feature auto-indexes Python, JS, and other library documentation with hybrid FTS5 search and version pinning. Ships with local Qwen3 embeddings so you can run it without API keys. Recent v2.0 removed media analysis in favor of the author's separate imagine-mcp server. Reach for this when you need reliable web scraping that doesn't choke on Cloudflare or when you want offline-capable doc search across multiple library versions.