What problem does it solve? Web search from servers often fails due to CAPTCHAs, datacenter IP blocks, and missing API keys. This Skill provides a resilient search workflow with a strict fallback chain so queries still return structured results when primary engines fail. ## Core Features & Use Cases - Four-Level Fallback Chain: SerpAPI/Brave/Serper APIs first, then a self-hosted SearXNG instance, then Startpage over Tor, then public engines as a last resort. - Unified Output Contract: Every result is normalized to {title, url, snippet, position, engine} regardless of the underlying engine. - Deployment Troubleshooting: Documented fixes for SearXNG 2026+ pitfalls (settings.yml vs .yaml, limiter.toml schema errors, msgspec missing, worker crashes) and IP-block diagnosis via Tailscale exit nodes or Tor SOCKS5. - Use Case: A research agent on a cloud server needs literature search results, but Google and Bing block its datacenter IP. The Skill diagnoses the block, routes traffic through a Tailscale exit node, and returns 10 normalized results. ## Quick Start Search the web for "S2 API paper retrieval" using the fallback chain and return 10 structured results with title, url, snippet, position, and engine.
