GrubCrawler is a self-hosted web crawler packaged as a Docker container that claims to bypass major bot detection systems (Cloudflare, Akamai) through fingerprint rotation and TLS cloaking, extract clean text and link structure from pages, and operate faster than commercial alternatives. The pitch emphasizes speed (concurrent fetching, connection pooling), stealth (detection bypass), and control—you run it yourself rather than depending on centralized AI lab infrastructure, with per-agent OAuth scoping. The source says it's open source on GitHub and available as a managed instance at grub.nuts.services, though the source doesn't provide actual benchmarks, performance numbers, detection rates, or technical depth on how the bypass mechanisms work. It reads as a product announcement rather than technical documentation or proof of claims.