Routes search queries across 14 providers with automatic failover and RRF fusion, then extracts full page content through a 12-step pipeline. Exposes search, URL extraction, dead URL recovery via Wayback Machine, site capture workflows, and research pack generation that bundles official docs with supporting sources. Starts with zero configuration using DuckDuckGo and Yahoo, scales up with API keys or self-hosted SearXNG for 70+ engines. Useful when you need reliable search and content extraction for RAG pipelines or agent workflows without stitching together multiple APIs. Includes topology-aware routing that adjusts behavior based on whether you're running on residential or datacenter IPs. Ships with a usage dashboard and persistent local corpus storage.
claude mcp add --transport stdio khamel83-argus uvx argus