# grok.com — bxc detect / recon (2026-06-03) Target: [https://grok.com/](https://grok.com/) Cookie jar: `~/.bxc/cookies/grok.json` (shortcut `grok`), mirrored to `~/.aphrody/cookies/grok.json`. ## Commands run ```bash # Without cookies (baseline) bxc detect https://grok.com/ --json bxc recon https://grok.com/ --profile stealth --timeout 60000 # With cookie jar (stealth browser) cd ~/bxc && bun run scripts/grok-scan-with-cookies.ts ``` ## Results summary | Probe | HTTP | Notes | | --- | --- | --- | | `bxc detect` (fetch) | **403** | Cloudflare; CDN = Cloudflare; NS = `*.ns.cloudflare.com` | | `bxc recon` (stealth, no cookie CLI flag) | **403** | Challenge page; CSP allows `challenges.cloudflare.com` | | Stealth browser + `cookies: "grok"` | **403** body | Title “Just a moment…” — CF managed challenge; clearance likely IP-bound or stale | Resolved IPs: `104.18.28.234`, `104.18.29.234` (Cloudflare anycast). ### Detect (CDN / DNS only) - **CDN:** Cloudflare (`cf-ray` header, IP range) - **DNS:** Cloudflare nameservers - **Frameworks:** none (body is challenge HTML, not app bundle) ### Recon assets (challenge shell) - Script: `/cdn-cgi/challenge-platform/h/g/orchestrate/chl_page/v1` - Selectors: `#challenge-error-text`, `.main-content`, `body`, `html` ## Artifacts (local, not committed) Full JSON and HTML snapshots live under: `docs/x/reports/grok-com/` (not in repo) (gitignored via `m3/docs/research/reports/`) Files: `detect.json`, `recon.json`, `browser-html-head.txt`, `browser-final-url.txt`, `README.md`. ## Recommendations 1. Refresh `cf_clearance` and `__cf_bm` from the same machine/IP as the VPS, or run `bxc cookies extract grok.com` on a desktop Chrome profile (if available). 2. For agent automation of Grok chat, use **Grok Build CLI** / **xAI API** — not bxc scrape of grok.com. 3. For X/Twitter automation, use `bxc x` / `aphrody-x` with `auth_token` + `ct0` (see [bxc-integration.md](./bxc-integration.md)).