Judgment-Free Consumer Money Coach

You are designing "[INSERT PRODUCT NAME]", a consumer AI money coach for [INSERT MARKET, e.g., first-generation middle-income households in Southeast Asia]. Output the full plan first; execute module by module after approval. 1. SCOPE: coaching and education, not discretionary advice; hand investments, insurance, and tax filing to licensed partners. 2. PERSONA: judgment-free, plain-spoken, encouraging; no shame language about debt or income. 3. FUNCTIONS: cash-flow tracking, budget envelopes, debt-payoff simulation, goal planning, bill-shock alerts, savings nudges. 4. DATA AGGREGATION: open-banking connections through [INSERT AGGREGATOR], transaction categorization, and manual accounts. 5. DATA FLOW: bank feeds to categorization to a rules engine to LLM explanations; user verification precedes any action. 6. DECISION SUPPORT: avalanche-versus-snowball comparisons, emergency-fund targets, and scenario simulations with assumptions shown. 7. INTERACTION: a weekly money-date ritual, one-tap category fixes, voice queries, and a partner-shared view with permissions. 8. NUDGE ETHICS: cap nudge frequency, reject dark patterns, and disclose partner commissions behind recommendations. 9. FRAUD AND SECURITY: read-only credentials, anomaly detection, scam-message warnings, and step-up auth for changes. 10. COMPLIANCE: local financial-advisor regulations, disclaimer copy, audit trails, and a per-country licensing map. 11. PRIVACY: never sell transaction data, encrypted vault, sensitivity flags for income and rent, deletion including backups. 12. MODEL RISK: deterministic calculations to prevent hallucinated figures, with every figure citing an actual transaction.