initializing

Detect & Refine

Detect & Refine watches every visitor — scroll depth, mouse velocity, rage clicks, dwell, exit intent — and grades the click A+ → F with rule-based scoring refined by Groq AI. Then it fans the offline conversion signal back to Google Ads, Meta, LinkedIn, TikTok, and X automatically, so the networks learn who to bid on (and who to stop showing ads to). 4 KB embed, sendBeacon, auto-captures gclid/fbclid/li_fat_id/ttclid/twclid. Per-tenant accounts with allowed_origins gating, monthly quotas, and audit-logged conversion sends. Ships as 15 tools inside 0nMCP — installable via `npm i 0nmcp` and configured through dr_account_create + dr_account_update. Or use the standalone REST surface at detect-and-refine-cryptogoatz.vercel.app.

What it does

  • 4 KB tracking script — non-blocking, async, sendBeacon
  • Auto-captures gclid, fbclid, li_fat_id, ttclid, twclid
  • 30+ behavioral signals: scroll, dwell, rage/dead clicks, form focus, tab visibility
  • Rule-based scoring refined by Groq llama-3.1-8b
  • Grades A+ → F (sale, nurture, dead, exclude)
  • Offline conversion fanout: Google Ads OCI, Meta CAPI, LinkedIn, TikTok, X
  • Per-tenant accounts with origin allowlist + monthly quota
  • Audit log of every network conversion send
  • Adaptive Siphon — live canvas of traffic quality
  • Install via 0nMCP (15 tools) or REST API

Capabilities required

behavioral_trackclick_grade_aiad_network_fanouttenant_accountsorigin_allowlistmonthly_quotasheet_exportrollup_stats

How it works

1

Install

Add Detect & Refine to your 0nCore account from the marketplace.

2

Connect

The add-on auto-connects to your CRM sub-location. No manual setup.

3

Use It

Access the capability from your dashboard, chat, or automations.

Integrations

Google AdsMeta AdsLinkedIn AdsTikTok AdsX AdsGoogle SheetsGroq AI0nMCP