Financial Advisor AI Agent

Write a production-ready system prompt for a financial advisor AI agent for [FIRM NAME], a [FIRM TYPE: RIA/broker-dealer/fintech app]. The agent provides personalized financial guidance. Requirements: 1) Role: a financial education assistant — NOT a licensed financial advisor, cannot give personalized investment advice without proper licensing disclosures, 2) Financial literacy calibration: on first interaction, assess the user's financial knowledge level (beginner/intermediate/advanced) and calibrate response depth accordingly, 3) Goal collection: understand their financial goals (retirement, home purchase, debt payoff, emergency fund, investing), timeline, and current situation, 4) Guidance approach: explain concepts clearly, use analogies for complex topics, always show the math, 5) Regulatory compliance: include appropriate disclaimers, never guarantee returns, never recommend specific securities without proper licensing, 6) Escalation: for complex situations (estate planning, tax optimization, business finances), recommend consulting a licensed professional, 7) Prohibited: never predict market movements, never guarantee investment returns. Include 3 few-shot examples: budgeting question, investment question, retirement planning question.

Build a Node.js TypeScript retirement projection calculator API. Endpoint: POST /api/retirement/projection. Input: { currentAge, retirementAge, currentSavings, monthlyContribution, expectedAnnualReturn, inflationRate, desiredMonthlyIncome }. Calculations: 1) Future value of current savings at retirement, 2) Future value of monthly contributions at retirement, 3) Total projected nest egg at retirement, 4) Monthly income this nest egg can sustain (4% rule and custom withdrawal rate), 5) Inflation-adjusted monthly income in today's dollars, 6) Savings gap: if projected income < desired income, how much more to save monthly, 7) Sensitivity analysis: show projections at return rates of (expected-2%), expected, (expected+2%), 8) Year-by-year projection table for charting (age, balance, contributions, growth). Return structured JSON with all calculations and a plain-English summary paragraph. Include input validation with Zod.

You are a financial planning AI assistant. Analyze the following budget and create a debt payoff plan. Monthly income (after tax): $[AMOUNT]. Monthly expenses: [LIST CATEGORIES AND AMOUNTS]. Debts: [LIST DEBTS WITH BALANCE, INTEREST RATE, MINIMUM PAYMENT]. Analysis: 1) Budget breakdown: categorize expenses as needs/wants/savings, calculate % of income for each, compare to 50/30/20 rule, 2) Cash flow: monthly surplus/deficit after all expenses and minimums, 3) Debt payoff strategies: compare Avalanche (highest interest first) vs. Snowball (smallest balance first) — show total interest paid and payoff timeline for each, 4) Recommended strategy: which approach is better for this specific situation and why, 5) Month-by-month payoff schedule for the recommended strategy, 6) Freed-up cash flow: what happens to the budget after each debt is paid off, 7) Quick wins: 3 specific expense reductions that would meaningfully accelerate payoff. Show all math. Use specific numbers, not percentages.