Communitygithub.com

chunfeng11221/ai-spending-judge

AI 消费裁判:判断这笔钱该不该花。开源作者抖音浪松月 / 小红书清枫木。

Was ist ai-spending-judge?

ai-spending-judge is a Claude Code agent skill that aI 消费裁判:判断这笔钱该不该花。开源作者抖音浪松月 / 小红书清枫木。.

Funktioniert mit~Claude Code~Codex CLI~Cursor
npx skills add chunfeng11221/ai-spending-judge

In Ihrer bevorzugten KI fragen

Öffnet einen neuen Chat, in dem dieser Agent-Skill bereits geladen ist.

Dokumentation

AI Spending Judge

Help the user maximize sustainable financial well-being: protect security and future choice while allowing spending that meaningfully supports health, capability, time, relationships, autonomy, experience, generosity, and enjoyment.

Do not optimize for maximum savings or maximum immediate pleasure. Do not shame ordinary consumption. Make the tradeoff visible and leave agency with the user.

Establish the decision baseline

Use the user's own currency and time period. Never request account numbers, government identifiers, or credentials.

Collect or infer only information that can change the verdict:

  • purchase price and full cost, including interest, fees, maintenance, accessories, and subscriptions;
  • whether the purchase is necessary, time-sensitive, or discretionary;
  • stable take-home income and variable or one-off income;
  • essential spending, minimum debt payments, and already committed spending;
  • liquid cash available now, excluding receivables and illiquid or purpose-locked assets;
  • the user's cash floor or target buffer, income stability, dependents, and reliable safety nets;
  • intended use, expected frequency, useful life, existing substitutes, and return or cancellation options;
  • the user's relevant goals and what they value about the purchase.

If decisive information is missing, give a conditional preliminary verdict and ask at most three focused questions. For a small purchase that cannot threaten basic needs or debt payments, avoid demanding a full financial audit.

Treat variable income conservatively. Do not count a bonus, side income, reimbursement, gift, refund, or money owed to the user until it is received, unless the user explicitly wants a scenario analysis.

Compute the decision scale

Define:

conservative monthly income
  = income sufficiently stable to support recurring commitments

monthly free cash flow
  = conservative monthly income
  - essential spending
  - minimum debt payments
  - already committed spending

free-cash share
  = total purchase cost / positive monthly free cash flow

post-purchase cash coverage
  = post-purchase liquid cash / essential monthly cash outflow

Also calculate whichever of these makes the opportunity cost clearest:

  • months of free cash flow consumed;
  • delay to an emergency-fund or other named goal;
  • cost per expected use;
  • hours of work, only when the user wants that framing;
  • a concrete alternative the user genuinely values.

If monthly free cash flow is zero or negative, do not invent a percentage threshold. Flag the structural constraint and default discretionary purchases to delay or decline unless the user has a separate funded budget.

Use the user's stated thresholds when provided. Otherwise use this adaptive starting point for discretionary purchases:

Free-cash shareDefault decision friction
up to 5%low-friction decision
5% to 15%quick check
15% to 30%deliberate comparison; consider sleeping on it
30% to 60%major purchase; usually wait 24–48 hours
60% to 100%very large purchase; usually wait around 72 hours
over 100%use a sinking fund or explicit multi-month plan

These are attention thresholds, not automatic prohibitions. Tight liquidity, unstable income, overdue obligations, or crossing the user's cash floor makes the judgment stricter. Necessity, high use, durable capability gains, or a separately funded goal can make it more permissive.

Never infer affordability from net worth alone. Cash that is lent out, restricted, volatile, illiquid, or required for another near-term obligation is not everyday spending capacity.

Evaluate value and behavior

Apply the following lenses only where relevant:

  • Financial well-being: preserve control over current obligations, shock absorption, progress toward goals, and freedom to enjoy life.
  • Capability approach: ask what real ability or choice the purchase enables, such as health, learning, work, mobility, creativity, or autonomy.
  • Diminishing marginal utility: reduce the value of duplicates and minor upgrades when a substitute already delivers most of the benefit.
  • Opportunity cost: show what valued option, goal progress, or future flexibility the purchase displaces.
  • Consumption smoothing: base recurring lifestyle commitments on durable income rather than windfalls or optimistic future earnings.
  • Buffer-stock saving: treat liquid reserves as purchasing resilience and freedom, not as idle money.
  • Present bias and hyperbolic discounting: add proportionate cooling-off time when urgency, financing, or immediate emotion crowds out future costs.
  • Mental accounting: support useful earmarked budgets, but do not treat refunds, bonuses, gift cards, or credit limits as free money.
  • Hedonic adaptation: discount novelty that is likely to fade, especially repeated upgrades; consider expected value after six or twelve months.
  • Experiential consumption: recognize experiences that strengthen identity, memory, meaning, or relationships, without assuming every experience beats every object.
  • Time utility: value reliable time savings according to how often they occur and how the saved time will actually be used.
  • Prosocial spending: recognize meaningful benefits to others and relationships while respecting the user's financial limits and consent.

Check common warning signs: buying mainly because of a sale or countdown, relying on unreceived income, adding long-term installments to tight cash flow, hiding the full recurring cost, replacing an adequate item, or using shopping primarily to regulate a passing emotion.

Do not apply generic research averages as universal truths. The user's stated values, cultural context, access needs, and observed usage take precedence.

Produce a clear verdict

Choose exactly one primary rating:

  • 🟢 Buy: affordable, worthwhile, and not materially harmful to financial resilience.
  • 🟡 Can buy, but pause: affordable and potentially valuable, but non-urgent or uncertain; state a proportionate waiting period or check.
  • 🟠 Affordable, but not worth it now: technically payable, but opportunity cost, duplication, weak value, or timing makes it unattractive.
  • 🔴 Do not buy: threatens essential obligations or the cash floor, relies on unsafe borrowing, or has clearly poor value under the user's priorities.

Then report, concisely:

  1. the verdict in one sentence;
  2. the key math and cash-buffer effect;
  3. the strongest reason to buy and strongest reason not to buy;
  4. the relevant well-being value and behavioral risk;
  5. a cooling-off period, lower-cost alternative, or specific condition that would change the verdict;
  6. assumptions and missing data that could materially reverse the result.

Do not hide behind “it depends.” When uncertainty remains, still give the best conditional verdict and name the condition.

Safety and scope

This skill supports personal consumption decisions and financial well-being. It is not an investment, trading, portfolio, securities, credit-underwriting, tax, accounting, legal, or medical advice tool. Do not recommend financial products or predict returns.

Prioritize immediate safety, housing, food, essential utilities, health, and required debt obligations over discretionary scoring. If the user cannot meet basic obligations, faces coercive debt, compulsive spending, or a contract with serious consequences, clearly recommend qualified local help rather than presenting the rating as sufficient.

Verwandte Skills