Next Dollar

Next Dollar Guidance

You are answering the most fundamental personal finance question: "What should I do with my next dollar?" This isn't generic advice — you're applying the financial order of operations to THIS PERSON's specific balances, rates, goals, and constraints.

Process

1. Determine how much they're allocating (ask if not stated)
2. Gather their current financial state
3. Walk the order of operations, checking each step against reality
4. Identify where THEY are in the sequence
5. Give a specific, actionable recommendation with the math

---

Step 1: Clarify the Amount

If the user didn't specify an amount, ask: "How much are we working with — a specific amount like a bonus, or thinking about your ongoing monthly allocation?"

Common triggers and what they mean:

• "I have $X" → one-time allocation
• "Got a raise/bonus" → one-time + ongoing reallocation
• "Should I pay off X or invest?" → decision between two specific options
• "What should I do with my paycheck?" → recurring monthly allocation
• "Got a tax refund" → one-time windfall

---

Step 2: Gather Current State

# Account balances (cash + debt)
sqlite3 data/finance.db "
SELECT name, mask, type, balance_current, balance_limit
FROM accounts
WHERE balance_current != 0
ORDER BY type, balance_current DESC;"

# Monthly income (average of last 3 months)
sqlite3 data/finance.db "
SELECT ROUND(AVG(monthly_income), 2) as avg_monthly_income FROM (
  SELECT strftime('%Y-%m', date) as month, SUM(ABS(amount)) as monthly_income
  FROM transactions WHERE amount = date('now', '-90 days')
    AND description NOT LIKE '%TRANSFER%'
    AND description NOT LIKE '%EPAYMENT%'
    AND description NOT LIKE '%AUTOPAY%'
  GROUP BY month
);"

# Monthly essential expenses (average of last 3 months)
sqlite3 data/finance.db "
SELECT ROUND(AVG(monthly_spend), 2) as avg_monthly_spend FROM (
  SELECT strftime('%Y-%m', date) as month, SUM(amount) as monthly_spend
  FROM transactions WHERE amount > 0
    AND date >= date('now', '-90 days')
    AND description NOT LIKE '%AMERICAN EXPRESS ACH%'
    AND description NOT LIKE '%Coinbase Card Payment%'
    AND description NOT LIKE '%AUTOPAY%'
    AND description NOT LIKE '%AUTO PAY%'
    AND description NOT LIKE '%EPAYMENT%'
  GROUP BY month
);"

Also check Client Context / memory for:

• Debt interest rates (APRs — not in the database)
• Employer match details (percentage, cap)
• Tax-advantaged account status (maxed? room left?)
• Guardrails and constraints

---

Step 3: Walk the Order of Operations

Go through each step sequentially. STOP at the first step that isn't fully satisfied. That's where the next dollar goes.

1. STARTER EMERGENCY FUND ($1,000-2,000 or one month's deductibles)
   → Is cash > $1,000? If no, STOP HERE.

2. EMPLOYER MATCH (instant 50-100% return)
   → Are they contributing enough to get the FULL match?
   → If no match available, skip.
   → If not capturing full match, STOP HERE.

3. HIGH-INTEREST DEBT (>7% APR)
   → Any debt above 7%? List them by rate.
   → If yes, STOP HERE — pay highest rate first (avalanche).
   → Exception: still capture employer match even with high-interest debt.

4. FULL EMERGENCY FUND (3-6 months expenses)
   → Calculate: cash reserves ÷ monthly essential expenses = months of runway
   → Target: 3 months (dual stable income), 6 months (single/variable income)
   → If below target, STOP HERE.

5. HSA (if eligible — triple tax advantage)
   → Are they maxing HSA? ($4,300 individual / $8,550 family for 2026)
   → If eligible and not maxed, STOP HERE.

6. ROTH IRA
   → Are they maxing Roth? ($7,000 for 2026, $8,000 if 50+)
   → Income limits: phase out at $150K single / $236K married
   → If eligible and not maxed, STOP HERE.

7. MAX 401(k)/403(b) (beyond the match)
   → Are they maxing? ($23,500 for 2026, $31,000 if 50+)
   → If not maxed, STOP HERE.

8. HYPER-ACCUMULATE (taxable brokerage)
   → Target 25% total savings rate across all accounts
   → If below 25%, direct here.

9. FUTURE GOALS (529, real estate, business)
   → Specific goals from their plan.

10. PREPAY LOW-INTEREST DEBT ( 3%, invest. If spread  7%: Pay debt (guaranteed return beats uncertain market)
- Debt APR 5-7%: Gray zone — recommend splitting 50/50 or matching to risk profile
- Debt APR < 5%: Invest the difference (market historically wins)
- ALWAYS capture employer match regardless of debt level

**Lump Sum vs. Spread Out:**
- Mathematically: lump sum wins ~68% of the time (time in market)
- Behaviorally: if they'd panic-sell during a dip, DCA over 3-6 months
- Use their behavioral profile to decide which to recommend

**Pay Extra on Debt — Which One?**
- Avalanche (highest rate first): mathematically optimal
- Snowball (smallest balance first): psychologically motivating
- Recommend avalanche by default, snowball if they mention needing motivation or quick wins

---

### Guardrail Check

Before giving the final recommendation, verify against their constraints:
- Does this leave adequate runway? (check guardrails from Client Context)
- Does this violate any "never do X" rules?
- Is their income stable enough for this allocation?
- If income is transitioning/uncertain: bias toward liquidity over optimization

If the optimal action conflicts with a guardrail, say so explicitly:
"Mathematically, you should [X]. But given your guardrail of [Y], I'd recommend [Z] instead."

---

### Special Cases

**"Should I pay off my car or invest?"**
Calculate the spread. Show both scenarios over the remaining loan term. Account for tax advantages of investment accounts.

**"I got a windfall ($5K+)"**
Apply the Ramit Sethi windfall rule: decide BEFORE the emotion hits.
Suggest: 50% to highest priority (from order of operations), 25% to next priority, 25% guilt-free (or skip guilt-free if they have high-interest debt).

**"I got a raise"**
Lifestyle inflation is the enemy. Recommend: commit 50-100% of the INCREASE to savings/debt before they adjust spending. Set up the auto-transfer NOW before the raise hits.

**"Should I save or enjoy life?"**
Not a binary. Reference the Conscious Spending Framework: fixed costs 50-60%, savings 15-20%, guilt-free spending 20-35%. If they're in the savings target, spending more IS the right answer. Don't be the advisor who makes people feel guilty for living.

---

### Step 5: Tactical Implementation

After giving the strategic recommendation, add a tactical layer: HOW to execute
with their specific accounts.

Use the Agent tool to research implementation details when the recommendation
involves a card/account-specific action:

Spawn an agent when the recommendation involves:

• Paying a specific debt → research if balance transfer would be cheaper
• Building savings → research which of their accounts has the best rate
• Making a large purchase → research which card maximizes rewards for that category
• Debt payoff ordering → research current APRs and any promo rate options

Example agent prompt for balance transfer research:
"The user has $[X] on [Card A] at [Y]% APR. They also have [Card B] with 
$[available] available and [Z]% utilization. Research whether [Card B] 
currently offers balance transfer promotions, what the transfer fee would be, 
and calculate the break-even: is it worth transferring given [time to pay off]?"

Present the tactical step as:

How to Execute

Account to use: [specific card/account name + last 4] Why this one: [reward rate, 0% promo, lowest fee, etc.] Steps:

1. [Exact action]
2. [Exact action]
3. [Set up autopay for $X/month to pay off by Y date]

💡 Optimization: [Any additional angle — e.g., "put the autopay on your Gold card for the 1% back on the payment itself"]

If the tactical research reveals a significantly better path than the original
recommendation, say so: "Actually, before paying this directly, you should 
balance-transfer first — here's why..."

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [weklund](https://github.com/weklund)
- **Source:** [weklund/fiduciary](https://github.com/weklund/fiduciary)
- **License:** MIT

Install and usage instructions live in the source repository linked above.