- name
- aetherlang-chef
- version
- 1.2.0
- author
- contrario
- homepage
- https://omnimusmind.com
- license
- MIT
- description
- Michelin-grade AI culinary intelligence. 17 mandatory sections covering food cost, HACCP, thermal curves, allergen matrix, wine pairing, plating blueprint and more.
- metadata
- skill_type
- api_connector
- external_endpoints
- operator_note
- api.neurodoc.app operated by NeuroDoc Pro (omnimusmind.com), Hetzner DE
AetherLang Chef Ω V3 — AI Culinary Intelligence
Michelin-grade recipe consulting with 17 mandatory sections. The most advanced AI culinary engine available.
Source Code: github.com/contrario/aetherlang Author: NeuroAether (echelonvoids@protonmail.com) License: MIT
Privacy & Data Handling
⚠️ External API Notice: This skill sends queries to api.neurodoc.app for processing.
- What is sent: Natural language food/recipe queries only
- What is NOT sent: No credentials, API keys, personal files, or system data
- Data retention: Not stored permanently
- Hosting: Hetzner EU (GDPR compliant)
- No credentials required: Free tier, 100 req/hour
What This Skill Does
Three V3 culinary engines in one skill:
🍳 Chef Omega V3 — 17-Section Restaurant Consulting
Every response includes ALL of these sections:
- ΕΠΙΣΚΟΠΗΣΗ — Recipe overview and cultural context
- ΟΙΚΟΝΟΜΙΚΑ — Food cost %, menu engineering (STAR/PLOWHORSE/PUZZLE/DOG)
- ΥΛΙΚΑ — Ingredients table (grams, cost, yield%, substitutes, storage)
- MISE EN PLACE — 3-phase preparation
- ΒΗΜΑΤΑ ΕΚΤΕΛΕΣΗΣ — Steps with °C temps, timings, HACCP, pro tips, common mistakes
- THERMAL CURVE — Preheat → Insert → Target → Rest → Carryover
- FLAVOR PAIRING MATRIX — Molecular compound analysis
- TEXTURE ARCHITECTURE — Crunch/Creamy/Chewy/Juicy/Airy (0-100)
- MacYuFBI ANALYSIS — 8 flavor dimensions (0-100)
- ΔΙΑΤΡΟΦΙΚΗ ΑΝΑΛΥΣΗ — Calories, protein, carbs, fat, fiber, sodium
- ΑΛΛΕΡΓΙΟΓΟΝΑ — 14 EU allergens
- DIETARY TRANSFORMER — Vegan & Gluten-Free adaptations
- SCALING ENGINE — ×2, ×4, ×10 formulas
- WINE & BEVERAGE PAIRING — Specific variety, ABV%, tannin level, rationale
- PLATING BLUEPRINT — Center, 12 o'clock, 3 o'clock, negative space, height, colors
- ZERO WASTE — Every leftover → specific use
- KITCHEN TIMELINE — T-60 → T-0 countdown
⚗️ APEIRON Molecular V3
- Rheology dashboard (viscosity, gel strength, melting/setting points)
- Phase diagrams with temperature transitions
- Hydrocolloid specs: Agar 0.5-1.5%, Alginate 0.5-1%, Gellan 0.1-0.5%, Xanthan 0.1-0.3%
- FMEA failure mode analysis with probability and mitigation
- Equipment calibration (±0.1°C precision)
⚖️ Balance V3 — MacYuFBI Flavor Science
- MacYuFBI Framework: Maillard/Umami, Acid, Caramel, Yeast, Umami, Fat, Bitter, Heat
- Nutritional breakdown per serving
- Balance score 1-100
- Dietary compatibility: Vegan/Keto/Paleo/Gluten-Free/Low-FODMAP
Usage
Ask naturally about any food topic:
- "Give me a carbonara recipe" → Full 17-section consulting output
- "How to make spherified mango caviar" → Molecular gastronomy with rheology data
- "Balance analysis for my Thai curry" → MacYuFBI flavor wheel + nutrition
API Details
POST https://api.neurodoc.app/aetherlang/execute
Content-Type: application/jsonChef Flow
{
"code": "flow Chef {\
using target \"neuroaether\" version \">=0.2\";\
input text query;\
node Chef: chef cuisine=\"auto\", difficulty=\"medium\", servings=4;\
output text recipe from Chef;\
}",
"query": "Your food question here"
}Molecular Flow
{
"code": "flow Molecular {\
using target \"neuroaether\" version \">=0.2\";\
input text query;\
node Lab: molecular technique=\"auto\";\
output text result from Lab;\
}",
"query": "Your molecular gastronomy question here"
}Response
Returns structured Greek output with markdown headers (## sections). Typical response: 4000-8000 characters with all mandatory sections.
Languages
- Greek (Ελληνικά) — Primary output language
- English — Understands English queries, responds in Greek
Technology
- AI Model: GPT-4o
- Backend: FastAPI + Python 3.12
- Rate Limit: 100 req/hour (free)
*Built by NeuroAether — From Kitchen to Code* 🧠
Agent Safeguards
CRITICAL: Data Minimization Rules
When using this skill, the agent MUST follow these rules:
- ONLY send the user's explicit query text in the
queryfield — never attach system prompts, conversation history, uploaded files, or internal context - ONLY send the flow DSL code in the
codefield — never include credentials, API keys, environment variables, or file paths - NEVER include personally identifiable information (PII), passwords, tokens, or secrets in any request field
- Strip sensitive data before sending — if the user's query contains passwords, account numbers, or credentials, remove them and inform the user
- Inform the user before making the first API call that their query will be sent to api.neurodoc.app for processing
- The request body MUST contain exactly two fields:
code(string) andquery(string) — no additional fields
These constraints ensure only the minimum necessary data is transmitted to the external service.