Serveur MCP — outils accessibles aux IA

Zoyku embarque un serveur MCP (Model Context Protocol) : les assistants IA comme Claude ou Cursor peuvent exécuter les mêmes calculateurs et convertisseurs que ce site — même logique testée, aucune réimplémentation.

Les outils exécutent les fonctions pures du site (conversion d'unités, finance, estimations santé, traitement de texte). Ils tournent via npx sur la machine du client IA — rien n'est envoyé à nos serveurs.

Ajouter à votre client MCP

Nécessite Node.js 18+. Le serveur tourne localement en stdio ; seuls les appels que vous lancez sont calculés sur votre machine.

{
  "mcpServers": {
    "zoyku": {
      "command": "npx",
      "args": [
        "-y",
        "zoyku-mcp"
      ]
    }
  }
}

Outils exécutables (35)

list_unit_categories	List all unit conversion categories available in convert_unit (length, weight, temperature, data, speed, volume, time, pressure, energy) with their unit ids and symbols.
convert_unit	Convert an amount between units. Get valid category/unit ids from list_unit_categories first. Examples: length m→ft, weight kg→lb, data GB→GiB, temperature c→f, pressure bar→psi, energy kcal→kj.
mortgage_payment	Monthly mortgage payment (P&I) via standard amortization, plus loan amount, total interest and total paid. Also returns the yearly amortization schedule.
compound_interest	Future value with regular contributions (annuity-due: contribution at period start). Returns total, contributed, interest and a year-by-year schedule.
salary_breakdown	Break an annual salary down to hourly/daily/weekly/biweekly/monthly. Optional unpaid days off raise the effective hourly rate (same salary over fewer worked days).
hourly_to_salary	Convert an hourly rate to daily/weekly/biweekly/monthly/annual pay.
break_even	Break-even units and revenue for fixed costs, unit price and variable unit cost; includes contribution margin and profit at 0.5×/1×/1.5×/2× of break-even volume.
bmr	BMR via the Mifflin-St Jeor equation (1990). Estimate only — not medical advice. Metric units.
tdee	Total daily energy expenditure = BMR (Mifflin-St Jeor) × activity factor (sedentary 1.2 … very active 1.9). Estimate only — not medical advice.
body_fat_navy	Body fat percentage via the US Navy circumference method (Hodgdon & Beckett 1984). Metric cm. Hip is required for female. Estimate only.
ideal_weight	Ideal weight by the Devine formula (1975, male and female) plus the healthy BMI 18.5–24.9 range for a given height. Metric cm.
water_intake	Baseline 35 ml/kg body weight plus ~350 ml per 30 minutes of exercise. Returns baseline, exercise and total ml. General guideline, not medical advice.
running_pace	Pace (per km and per mile) and speed from a race distance and finish time. Distance: '5k', '10k', 'half', 'marathon' or custom meters.
race_time_predict	Predict a race time from a recent result using Riegel's formula (exponent 1.06). Distances in km (5, 10, 21.0975, 42.195 or custom).
one_rep_max	Estimated 1RM from weight lifted and reps (Epley and Brzycki formulas).
workdays_between	Count working days (Mon–Fri by default) between two ISO dates, inclusive. Optional holidays (YYYY-MM-DD).
add_workdays	Add N working days to an ISO date (negative to subtract). Optional holidays. Returns the resulting ISO date.
sleep_schedule	Bedtimes to wake at a given time, or wake times if sleeping now — based on 90-minute sleep cycles plus 15 minutes to fall asleep. Cycles vary 80–120 min between individuals.
cooking_convert	Convert between volume and weight for ~24 ingredients (flour, sugar, butter, honey…) using average densities. Units: cup, tbsp, tsp, fl-oz, ml, l, g, kg, oz, lb. Densities are averages — weigh for baking precision.
contrast_ratio	Contrast ratio between two colors (#RGB/#RRGGBB) with WCAG AA/AAA pass-fail for normal and large text.
grade	Weighted grade from scored items, plus letter grade on a configurable scale.
slugify	Turn text into a URL-safe slug (lowercase, accents stripped, hyphens).
text_stats	Word count (CJK-aware), characters with/without spaces, sentences, paragraphs, lines.
transform_text	Case and cleanup transforms: title_case, camel, pascal, snake, kebab, dedupe_lines, sort_lines, reverse, clean (collapse spaces/trim), remove_extra_spaces.
parse_csv	Parse RFC-4180 CSV (quoted fields, embedded newlines, ,/;/tab delimiter auto-detect) into an array of row objects using the first row as header.
json_to_xml	Convert a JSON object to XML. Arrays become repeated elements; '@'-prefixed keys become attributes; '#text' becomes text content. Mixed content is not representable.
overtime_pay	Weekly and annualized pay with overtime: hourly rate, overtime hours and a multiplier (US FLSA standard is 1.5× over 40 h/week).
rent_affordability	Maximum rent by an income-share rule (default 30% of monthly income), with debt payments deducted and the resulting debt-to-income share reported.
apy	Effective annual yield (APY) from a nominal rate and compounding frequency: APY = (1 + r/n)^n − 1. Also returns the reverse (APY → nominal APR).
gpa	GPA from courses with credits and letter grades (A…F quality points). Invalid courses are dropped and reported.
heart_rate_zones	Five training zones via Karvonen's heart-rate-reserve method. Max HR defaults to the Tanaka formula (206.9 − 0.67 × age). BPM ranges per zone.
shoe_size	Convert an adult shoe size between US/UK/EU/JP systems for men or women, anchored to foot length in cm. Brands vary ±half size.
json_to_typescript	Generate TypeScript interfaces from a JSON object: nested objects become named interfaces, arrays of objects merge into one shape, mixed arrays become unions.
json_diff	Compare two JSON documents path by path: added, removed and changed keys with full paths like $.user.name. Arrays compared by index.
json_query	Extract a value from JSON by path: $.a.b, $.items[0].id, $['quoted key']. Returns found/value or a plain-language reason.

Nécessite Node.js 18+. Le serveur tourne localement en stdio ; seuls les appels que vous lancez sont calculés sur votre machine.