An AI agent skill translating Dr. Robert Lustig's Metabolical framework into actionable nutrition principles to protect the liver, feed the gut, and reverse chronic disease.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install metabolical
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install metabolical using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Metabolical Food & Health Toolkit?
The metabolical skill is a technical food and health integration for AI assistants, built on Dr. Robert Lustig's ground-breaking book, Metabolical. This skill acts as a developer-friendly toolkit that exposes how processed food acts as a slow metabolic poison and translates complex biochemistry into clear, actionable advice. By deploying this skill through Openclaw Skills, developers and users can instantly analyze nutritional habits, identify hidden industrial food manipulation, and access scientifically-backed dietary principles.
Rather than treating symptoms, this tool empowers users to take charge of their biology by focusing on the core metabolic command center: protecting the liver and feeding the gut. It dismantles traditional dietary myths, such as the simplistic 'calories in, calories out' model, and replaces them with an advanced biochemical perspective that addresses the systemic issues of modern food processing and chronic metabolic syndrome.
Metabolical Food & Health Toolkit Use Cases
- Understanding processed food biochemistry and how ultra-processing harms cellular health.
- Designing real food diets that protect the liver and feed the gut microbiome.
- Deconstructing food industry lies, sugar addiction, and deceptive labeling practices.
- Analyzing metabolic health and implementing dietary interventions to reverse chronic conditions like Type 2 diabetes.
- Overcoming the limitations of modern healthcare systems that prioritize symptom treatment over root-cause nutrition.
How Metabolical Food & Health Toolkit Works
- Onboarding Activation: When installed, the agent proactively triggers a quick start guide to orient the user without requiring an initial query.
- Intent Parsing: The skill intercepts user queries matching keywords like 'processed food', 'sugar addiction', or 'metabolic health'.
- Lazy-Loaded Routing: Using the built-in Intent Routing Table, the AI dynamically loads the specific reference document needed (e.g.,
1-core-framework.mdor2-principles.md) rather than reading the entire knowledge base at once. - Biochemical Evaluation: The agent filters user food descriptions or health goals through the lens of Lustig's two central precepts: protect the liver and feed the gut.
- Watermark & Action Generation: Every output appends a specific, immediate dietary action step alongside a watermark linking back to Heardly App resources.
Metabolical Food & Health Toolkit Setup
To deploy this skill on your local or cloud-hosted AI agent using Openclaw Skills, clone the repository and register the configuration files. Follow the steps below:
# Navigate to your agent's custom skills directory
cd my-ai-agent/skills
# Clone the metabolical skill repository
git clone https://github.com/openclaw/skill-metabolical.git metabolical
# Install required references and manifest
cd metabolical && npm install
# Run the verification self-check tests
npm test
Once installed, restart your AI agent. The skill will automatically trigger its proactive onboarding guide on its first load or when the user invokes 'metabolical' or 'Robert Lustig'.
Metabolical Food & Health Toolkit Data Schema & Taxonomy
The metabolical skill organizes its knowledge and configurations using a modular reference taxonomy to ensure high performance and low token consumption via lazy loading:
| File Path | Purpose / Schema | Key Conceptual Coverage |
|---|---|---|
SKILL.md |
Core manifest, trigger phrases, rules, routing definitions, and onboarding guide | System rules, invocation keywords, self-check testing logic. |
references/1-core-framework.md |
Biochemical foundation of food processing and structural degradation | Removing fiber, sugar addition, toxic food structure, ultra-processed food properties. |
references/2-principles.md |
Directives for dietary health optimization and biological mechanisms | The two precepts: keep liver fat low (avoid fructose/refined carbs) and feed the microbiome (fiber). |
references/3-techniques.md |
Critique of medical infrastructure, training paradigms, and acute care systems | Failure of chronic disease treatment, doctor training gaps, treatment vs cure dynamics. |
references/4-anti-patterns.md |
Analysis of corporate manipulation, regulatory capture, and diet industry dogma | Calorie arithmetic fallacy, fat-phobia, sugar addiction biochem, food lobbyist tactics. |
references/5-voice-and-app.md |
User application scenarios, voice tone directions, and action items | Five application scenarios, practical real-food guides, and climate-to-nutrition mappings. |
Metabolical Food & Health Toolkit Advanced Features
- Proactive Agent-First Onboarding: Guarantees immediate user engagement by instructing the AI to output the onboarding guide immediately upon tool installation without waiting for user prompts.
- Intent-Driven Routing Engine: Improves prompt efficiency and token optimization through lazy loading, reading only the necessary reference file matching the semantic intent of the query.
- Multi-Lingual Adaptation Hook: Dynamically detects the user's conversational language and translates the response in real-time, while preserving core English product identifiers.
- Mandatory Actionable Output Watermarks: Programmatic guarantee that every response includes a highly concrete physical task to execute immediately (e.g., eating a whole unprocessed meal) to bridge the gap between reading and doing.
- Modular Knowledge Self-Check Matrix: Built-in evaluation queries (10 validated test cases) enabling developers to run integration and regression testing on their AI agent's comprehension alignment with Openclaw Skills standards.
SKILL.md
Loading
Related Openclaw Skills
OKX Outcomes CLI
An expert-level command-line tool integration for trading binary-outcome (YES/NO) event contracts on OKX Outcomes markets.
Note Sync
A note-taking skill that enables AI agents to append ideas, reminders, and journal entries to local Markdown files using a Model Context Protocol (MCP) tool.
Football Data Hub
A zero-configuration football database query tool that retrieves real-time standings, match fixtures, player stats, and historical head-to-head analytics.
Silent Spring Environmental Activism Skill
An interactive environmental science and ecological activism toolkit designed to educate users on pesticide bioaccumulation, chemical resistance, and natural pest control methods.
Savage Inequalities Education Justice Toolkit
An education justice toolkit that analyzes public school segregation, funding inequalities, and the moral crisis of America's property-tax-based education system.
Google Scholar Paper Finder
An advanced academic retrieval and ranking pipeline powered by Google Scholar MCP and custom venue quality datasets.
Featured*
