feat: add fastCRW search tool

[us]

What

Adds a fastCRW search tool (CRWSearch), alongside FirecrawlSearch.

New crw_search.py, registered alongside Brave/Bing/Google.

Why

fastCRW is a fully open-source web-scraping engine (AGPL, single ~8 MB Rust binary) that outperforms Firecrawl on Firecrawl's own benchmark dataset and runs entirely locally — no cloud dependency.

More open — actually works on protected/JS-heavy sites when self-hosted. Anti-bot/stealth bypass, BYO-proxy + rotation, and JS rendering all ship in the open core. Firecrawl's OSS edition gates its stealth engine (fire-engine) behind a cloud-only flag, so its self-host silently falls back to plain fetch/Playwright and can't reach Cloudflare-protected or JS-heavy pages. fastCRW handles Cloudflare JS challenges, UA rotation, SPA rendering, and an HTTP→headless→proxy fallback ladder — one binary, no multi-service stack, no cloud required.

Faster and higher recall on Firecrawl's own benchmark dataset. Truth-recall 63.74 % vs 56.04 %; median latency p50 ~1.9 s vs ~2.3 s. RAM footprint ~6 MB.

Search is built on SearXNG, not an alternative to it. SearXNG is the metasearch aggregator underneath; crw adds a quality layer on top: query expansion (multi-variant rewrite), content-aware reranking (re-scoring by fetched content instead of SearXNG's content-blind ordering), and category routing (research queries fan out to arxiv / semantic scholar / google scholar, code queries to GitHub). So you get SearXNG's breadth plus a measurable accuracy layer — all open-source (AGPL) and self-hostable.

Flat pricing: 1 credit = 1 page, no 4× stealth surcharge, no billed-on-failure. Key via CRW_API_KEY (free tier at https://fastcrw.com/dashboard); self-host base URL supported.

Firecrawl-API compatibility is the reason this integration is a tiny additive diff — FirecrawlSearch is untouched. Happy to adjust to project conventions; I maintain fastCRW and can provide free credits for testing.