This repository is the AI Operating System for a Gut Health & Nutrition Influencer Marketing Agency.
Its purpose is to provide a structured, maintainable and extensible knowledge base that enables AI agents and the human operator to collaboratively build, operate and improve the agency.
This repository is intended to become the single source of truth for the business.
The repository is built around three principles:
- Understand before building.
- Document before automating.
- Verify before completing.
AI agents should always prefer evidence over assumptions and reusable systems over temporary solutions.
Gut Health IMA/
|
├── bootstrap/
├── business/
├── source-material/
└── generated/
Contains the onboarding documents for AI agents.
These files explain:
- the mission of the project;
- the execution plan;
- the operating principles.
Every AI agent should read the required bootstrap rules first.
Contains authoritative information specific to this agency.
Examples include:
- business objectives;
- niche;
- services;
- positioning;
- operator decisions.
This folder represents the current state of the business.
Treat this folder as a protected input unless the operator explicitly authorizes a change.
Contains the authoritative training material used to build the agency.
These documents define the methodology.
They are considered read-only.
Do not modify these files.
Generated documentation must always remain traceable back to this material.
Contains AI-generated work.
Examples include:
- orientation reports;
- methodology maps;
- architecture proposals;
- operating manuals;
- workflows;
- agent definitions;
- research;
- reports.
Unless explicitly instructed otherwise, new work should be created here.
For initial repository orientation, a new AI agent should understand the repository in this order:
bootstrap/business/source-material/
For normal task execution, agents should read the required bootstrap rules first, then load only the business/ and source-material/ files relevant to the current task.
Do not begin implementation until the relevant authoritative context has been read and understood.
AI agents should:
- understand before implementing;
- ask when important information is missing;
- separate facts from assumptions;
- preserve traceability;
- avoid duplicating knowledge;
- build incrementally;
- verify work before considering it complete.
AI agents should not:
- modify protected inputs without explicit operator authorization;
- invent methodology;
- overwrite business decisions;
- create unnecessary complexity;
- assume previous conversation history exists.
The repository has completed its approved methodology map and initial runtime foundation.
The first runtime slice has been implemented and passed validation. The repository now has the minimum operating rules, planning skills, validation standards, and auditing behaviour required to support supervised business-capability development.
The current phase is Supervised Business-Capability Development.
Current priorities are:
- use the operating system rather than expand it prematurely;
- preserve protected authoritative inputs;
- build business capabilities incrementally from approved source material and operator decisions;
- verify each capability before considering it complete.
The next planned capability is Influencer Discovery.
The repository will be considered successful when a newly onboarded AI agent can:
- understand the purpose of the business;
- understand the operating methodology;
- understand the repository structure;
- identify authoritative sources;
- contribute productively without relying on previous chat history.
The repository-not conversation history-should become the long-term memory of the project.