growth-loops

[SKILL] growth-loops [OBJ] Engineer compounding acquisition loops — viral mechanics, referral programs, product-led growth funnels, and retention/re-activation systems. Each user should bring ≥1 new user (k-factor ≥ 1). Free-first: the viral unit must be usable without payment. RTL + Arabic parity for referral copy, share cards, and reward redemption. [RULES]

1. [REQ] Never invent metrics — read real funnel data from analytics (PostHog/Mixpanel) before proposing loops.
2. [REQ] Quantify the loop: k-factor, activation rate, D7/D30 retention, referral conversion rate. Present as a table.
3. [REQ] Prefer double-sided rewards (inviter + invitee) to maximize symmetric incentive. Calculate break-even CAC vs paid acquisition.
4. [REQ] Map the full loop: activation → retention → referral → revenue. Identify the weakest node and propose a fix.
5. [REQ] Pick the highest-leverage loop per stage: invite (K-factor > 1), share (content virality), embed (product distribution), credit (referral economy), content (UGC flywheel).
6. [CMD] Reuse runtime.attribution_model and runtime.funnel_tracker for measurement. Pull data via analytics_tools MCP (GA4/Mixpanel).
7. [REQ] Document the loop as a workflow step when it spans >1 tool. Store loop metrics via aizee memory add.
8. [REQ] Test loop viability: simulate k-factor from historical referral data before shipping. Gate behind aizee check for any automated outreach.

[PROHIBIT]

1. No fake "limited invite" false scarcity or dark patterns.
2. No forced-contact scraping or unauthorized address-book uploads.
3. No paid-placement claims presented as organic virality.
4. No bypass of approval/guardian gates for growth automations.