# No-Buy Pact > A planner for personal no-buy and low-buy challenges - the spending-reset trend where people stop buying non-essentials for a set period, keep a few deliberate allowances and check in on themselves. Sort sixteen spending categories as a swipe deck, see the month before and after, get a written pact, track clean days, and test any temptation against it. Not financial advice. URL: https://no-buy-pact.skillsafe.ai/ API: https://no-buy-pact.skillsafe.ai/api.html ## What runs free in the browser (packkit.js, deck.js) - The deck: 16 categories (clothes, takeout, coffee, restaurants, beauty, home, tech, online marketplaces, books/games, subscriptions, hobby, going out, gifts, and three essentials - groceries, bills and housing, transport and health). Drag left = cut, right = keep, up = limit with a monthly cap; buttons and arrow keys do the same; undo. - Amounts: typical monthly estimates (rough, for pre-filling), blank, or a bank CSV - the browser finds the description and amount (or debit) columns, skips income, transfers, card payments and refunds, buckets merchants by keyword and averages per month. The file never leaves the browser. - The receipt: spending now, after the pact, freed up per month (cut = whole amount, limit = amount above the cap, an essential marked cut = nothing) and over the challenge (months of 30.44 days). These are estimates from the card amounts. - Flags: essential cut, cap at or above current spending, everything cut, nothing cut, long and strict, unsorted cards, no amounts. - Starter rules: one plain rule per cut or limited card. - Tracker: one square per challenge day; tap for clean, again for slip; current and longest streak, unlogged days, days left. - Cap check for a temptation: category status, allowance left, how far over, hours of work at your take-home rate, cost per use, the date a 72-hour wait ends, and a rule-based hint (skip / wait / fits). Essential categories are never "skip", and the check says a real need (medicine, care, food, a repair) is bought now - the wait is for wants. - Temptation log: every logged temptation stays open until you mark it skipped or bought; a 72-hour wait that has run out is flagged as due, and skipped prices add up to money not spent. The calendar check-in links back to the tracker, and a returning visit opens your latest saved pact. - Exports: Markdown, CSV, calendar (.ics) of check-ins, JSON, a printable pact with a tracker grid. History saved to your account (collection `pacts`) with a browser mirror. ## What the model does (metered, model gpt-terra) - task "pact" (built on @joellewis/finance-psychology and @jmsktm/habit-tracker): a title, a first-person pledge, rules (a hard rule per cut category, an allowance per limit with the exact cap, 0-3 narrow exceptions), 2-5 triggers each with the cognitive bias behind it (present bias, loss aversion, mental accounting, scarcity, social proof, moral licensing and others from a fixed list) and a counter-move, 3-6 swaps, check-ins (daily for 30 days or fewer, else weekly; three questions; a streak rule), a slip plan, warnings for every strictness flag. - task "tempted": a fits / wait / skip verdict on one purchase, the pact rules that apply, the bias in the stated reason, three questions, a wait plan, a swap, and how to buy within the pact. - "Test a temptation against this pact" sends the pact's rules with the next temptation. ## What the page checks in every reply Pact: a cut card without a hard rule, a limit whose cap changed, a keep card that got banned, an essential banned outright, an unknown category, a bias outside the list, warning flags with no warnings. Temptation: "fits" on a cut category with no exception rule, "fits" over the remaining allowance, "skip" on an unrestricted category (marked as advice), a wait under 24 hours. ## Sources Derived from the SkillSafe skills @joellewis/finance-psychology (https://skillsafe.ai/skill/@joellewis/finance-psychology) and @jmsktm/habit-tracker (https://skillsafe.ai/skill/@jmsktm/habit-tracker).