AI Orchestrator
DE Early Access

AI Founder Operating System

Run your project.
Orchestrate your AI team.

Stop rebuilding your AI workflow from scratch. A file-based operating system for founders working with ChatGPT, Claude, Codex and AI agents — built from running a real production project.

You are not buying the models — you are buying the operating system around them: context, decisions, roles, routing, quality gates, handoffs and the next best action, across sessions and models.

Explore the operating system Get early access

Example role assignment — fully configurable, not a fixed ranking. Available models change over time. The system records these decisions; it does not access your AI accounts or monitor usage.

The problem

Founders don’t just lose time writing prompts.

They lose context, decisions, task history, quality standards and coordination between different AI tools.

  • Context is lost between sessions.
  • Decisions stay buried in chat history.
  • The same tasks get done twice.
  • Agents work from an outdated state.
  • AI tools overwrite each other’s work.
  • No one knows which model should take which task.
  • Expensive models are wasted on small tasks.
  • QA and handoffs get forgotten.
  • The founder builds setup instead of product.

The system

Two connected layers, one operating system.

One layer runs the project. The other coordinates the AI team. The operating system keeps them in sync.

Project System

Organizes the project.

FOUNDER_DECISIONS.md

PROJECT_STATUS.md

Shared State

Both layers read and write the same current state.

  • synchronized
  • current
  • verified
  • versioned

AI Team System

Coordinates the AI team.

AI_ROLE_MATRIX.md

HANDOFF_LOG.md

AI Founder Operating System One shared operational state for the project, decisions and AI execution.

Select a module to see its connection through the shared state.

Activate modules with mouse or keyboard — the matching connection lights up.

Three products

Choose a layer — or run both as one.

In development

AI Founder Kit

Organizes the project.

For
Solo founders and small teams running a project with AI assistance.
Problem
Context, decisions and task history live in scattered chats and get lost between sessions.

Core modules

  • Human Quick Start
  • AI Start File
  • Project Context
  • Founder Decisions
  • Task System
  • Handoffs
In development

AI Orchestrator Kit

Coordinates the AI team.

For
Founders working across several AI tools and reasoning modes.
Problem
No clear rules for which model does what, how work is handed off, or when an agent must stop.

Core modules

  • AI Role Matrix
  • Model Routing
  • Mode Selection
  • Task Routing
  • Repo Gates
  • Worktrees
In development Complete system

AI Founder Operating System

Keeps both in sync.

For
Founders who want the project and the AI team to stay in sync.
Problem
Project organization and AI coordination drift apart, so execution and quality suffer.

Real workflow

One idea, routed end to end.

Verified

  • Project status updated
  • Handoff recorded
  • Next best action ready
  • Strategy ChatGPT
  • Implementation Claude
  • QA Codex
One system, many specialized arms: different AI roles work independently — connected by shared context, rules and verification.
  1. Founder idea Reconciled with recorded context.
  2. ChatGPT strategy Direction and priority.
  3. Scoped task Routable and reviewable.
  4. Claude / Codex Executed within repo rules.
  5. QA Reviewed and verified.
  6. Commit A known-good state.
  7. Handoff Result preserved.
  8. Next best action Clear continuation.
Artifacts
  • Decision
  • Task
  • QA report
  • Handoff
  • Project status
  • Verified
A representative path from idea to the next best action.

What you actually get

One package. One connected system.

Not a loose prompt collection: project context, routing, verification and handoffs work together as one operating workflow.

This is not a prompt pack — it is one connected operating system of project, routing, QA and handoff modules.

Activate a module — connections and short infos appear.

  • PROJECT_CONTEXT.md

    Project Layer

    Keeps the project goal, context and assumptions stable.

    View excerpt Close excerpt
    Excerpt
    Project
    FeedbackPing — a lightweight feedback tool for appointment-based businesses
    Current phase
    Pre-launch — the first real-world test is next

    Included in: Founder Kit + Bundle

  • ACTIVE_TASKS.md

    Project Layer

    Shows what is being worked on right now and what comes next.

    View excerpt Close excerpt
    Excerpt
    Task
    TASK-001 — add the comment field, scoped to allowed files
    Protected
    The data schema only changes with explicit approval

    Included in: Founder Kit + Bundle

  • MODEL_ROUTING.md

    AI Team Layer

    Assigns tasks to the models and modes that fit them.

    View excerpt Close excerpt
    Excerpt
    Strategy / sparring
    ChatGPT — direction and priorities
    Implementation
    Claude — scoped to the rules of the project

    Included in: Orchestrator Kit + Bundle

  • QUALITY_GATES.md

    AI Team Layer

    Secures reviews and quality checks before the next step.

    View excerpt Close excerpt
    Excerpt
    Gate
    QA 7/7 passed — reviewed by a different role
    Rule
    Nothing counts as done without passing gates

    Included in: Founder Kit + Bundle

  • HANDOFF_LOG.md

    Shared State

    Preserves results and handoffs across sessions and models.

    View excerpt Close excerpt
    Excerpt
    Handoff
    TASK-001 completed — recorded on agent and project level
    State
    The verified state replaces the old one — nothing gets lost

    Included in: Founder Kit + Bundle

Illustration: a dark system vault with tentacles reaching out to glowing connector nodes — the product modules dock onto these strands.
  • PROJECT_CONTEXT.md Project Layer

    Keeps the project goal, context and assumptions stable.

    Excerpt
    Project
    FeedbackPing — a lightweight feedback tool for appointment-based businesses
    Current phase
    Pre-launch — the first real-world test is next

    Included in: Founder Kit + Bundle

  • ACTIVE_TASKS.md Project Layer

    Shows what is being worked on right now and what comes next.

    Excerpt
    Task
    TASK-001 — add the comment field, scoped to allowed files
    Protected
    The data schema only changes with explicit approval

    Included in: Founder Kit + Bundle

  • MODEL_ROUTING.md AI Team Layer

    Assigns tasks to the models and modes that fit them.

    Excerpt
    Strategy / sparring
    ChatGPT — direction and priorities
    Implementation
    Claude — scoped to the rules of the project

    Included in: Orchestrator Kit + Bundle

  • QUALITY_GATES.md AI Team Layer

    Secures reviews and quality checks before the next step.

    Excerpt
    Gate
    QA 7/7 passed — reviewed by a different role
    Rule
    Nothing counts as done without passing gates

    Included in: Founder Kit + Bundle

  • HANDOFF_LOG.md Shared State

    Preserves results and handoffs across sessions and models.

    Excerpt
    Handoff
    TASK-001 completed — recorded on agent and project level
    State
    The verified state replaces the old one — nothing gets lost

    Included in: Founder Kit + Bundle

Shared operating state

  • Founder Kit runs the project
  • Orchestrator Kit coordinates the AI team
  • INTEGRATION_FLOW.md ties both layers together
  • SAMPLE_PROJECT shows the whole loop once, end to end
  • AI Founder Operating System One operating workflow instead of four loose parts.

Fictional sample project — no real people or companies. Real file excerpts on the product pages: Founder Kit Orchestrator Kit Operating System

What is actually inside

Real files, not a black box.

Each kit is a set of working files you add to your project. These are the actual filenames that ship in v1.0 — no screenshots of an app that doesn’t exist.

Project layer

AI Founder Kit Organizes the project.

Entry points

  • START_HERE_FOR_HUMAN.md Where the project stands and the next best action.
  • START_HERE_FOR_AI.md The first file an agent reads — context, rules, priorities.

Project memory

  • PROJECT_CONTEXT.md What/who/why and the constraints that frame decisions.
  • PRODUCT_STRATEGY.md Positioning, scope and priorities.
  • FOUNDER_DECISIONS.md Append-only decision memory, kept out of chat history.

Working state

  • PROJECT_STATUS.md The verified current state.
  • ACTIVE_TASKS.md Scoped tasks in flight, with allowed files and a definition of done.
  • ROADMAP.md Phased plan of now / next / later.
  • NEXT_BEST_ACTION.md A single, unambiguous next step.

Guard rails & history

  • QUALITY_GATES.md The bar a change must clear to be "done".
  • STOP_RULES.md Hard conditions under which work pauses instead of guessing.
  • HANDOFF_LOG.md Append-only timeline of completed work.
  • templates/ Task, handoff, decision, status, roadmap and QA report formats.

AI team layer

AI Orchestrator Kit Coordinates the AI team.

Entry points

  • START_HERE_FOR_HUMAN.md Human orientation for the coordination layer.
  • START_HERE_FOR_AI.md Binding instructions an agent reads before acting.

Roles & routing

  • AI_ROLE_MATRIX.md Which AI takes which role, with separation of duties.
  • MODEL_ROUTING.md How to choose a tool and model for a task.
  • MODE_SELECTION.md The smallest reasoning mode that safely solves it.
  • TASK_ROUTING.md How a raw task is shaped, scoped and sent.
  • RESOURCE_POLICY.md Deliberate use of expensive modes.

Repository safety

  • REPO_GATE.md Pre-change checks that protect the repository.
  • WORKTREE_POLICY.md Isolating parallel work so agents do not collide.
  • CONFLICT_PREVENTION.md Keeping parallel agents off the same files.

Coordination & history

  • HANDOFF_PROTOCOL.md How one agent passes verified work to the next.
  • REVIEW_AND_VERIFICATION.md How results are reviewed and independently verified.
  • ESCALATION_RULES.md When to escalate to a stronger model or a human.
  • STOP_RULES.md When an agent must stop rather than continue.
  • TASK_HISTORY.md Append-only trail of routed tasks and results.
  • templates/ Implementation, research, review, QA, handoff and stop-report formats.

Shared state

AI Founder Operating System Keeps both in sync.

Bundle layer

  • SHARED_FILE_MAP.md Single source of truth for every shared concept across both kits.
  • INTEGRATION_FLOW.md The end-to-end loop that ties project and AI team into one motion.
  • INSTALLATION.md One installation for both kits into the same project.

Worked example

  • examples/sample-project/ A small, fully filled-in project showing the whole loop once.
  • PROJECT_CONTEXT.md A filled-in context for the fictional sample project.
  • MODEL_ROUTING_DECISION.md One real model-and-mode routing decision.
  • QA_REPORT.md One QA report with all gates passing.

Product preview

Real excerpts, not promises.

What the files look like once they are filled in — limited excerpts from a fictional sample project, one on every product page.

PROJECT_CONTEXT.md Excerpt

Project
FeedbackPing — a lightweight feedback tool for appointment-based businesses
Primary goal
Ask one good question after each appointment and see the answers in one place
Current stage
Pre-launch — the first real-world test is next
Users
Small service businesses, e.g. a two-person bike repair shop
Constraints
One founder, a few evenings per week, running cost near zero
Protected areas
The data schema changes only with explicit approval
Definition of success
Setup in under five minutes; at least half of the prompts get an answer

Fictional sample project — no real people or companies.

The project context survives outside individual chats — every session and every AI tool starts from the same state.

All product previews: Founder Kit Orchestrator Kit Operating System

Compare

One layer, the other, or both as one.

The Founder Kit runs the project. The Orchestrator Kit runs the AI team. The Operating System is the complete system that keeps both in sync.

Capability Founder Kit Orchestrator Kit Operating System Complete system
Project context
Decisions and roadmap
Tasks and handoffs
Quality gates
AI role matrix
Model routing
Repo / worktree safety
Review and verification
Integrated end-to-end loop Bundle only
Sample project Bundle only
Best for Running the project Coordinating the AI team Both, kept in sync
Current status Pre-release Pre-release Pre-release
Founder Kit
Project context
Included
Decisions and roadmap
Included
Tasks and handoffs
Included
Quality gates
Included
AI role matrix
Not included
Model routing
Not included
Repo / worktree safety
Not included
Review and verification
Not included
Integrated end-to-end loop
Not included
Sample project
Not included
Best for
Running the project
Current status
Pre-release
Orchestrator Kit
Project context
Not included
Decisions and roadmap
Not included
Tasks and handoffs
Shared
Quality gates
Partial
AI role matrix
Included
Model routing
Included
Repo / worktree safety
Included
Review and verification
Included
Integrated end-to-end loop
Not included
Sample project
Not included
Best for
Coordinating the AI team
Current status
Pre-release
Complete system Operating System
Project context
Included
Decisions and roadmap
Included
Tasks and handoffs
Included
Quality gates
Included
AI role matrix
Included
Model routing
Included
Repo / worktree safety
Included
Review and verification
Included
Integrated end-to-end loop
Included
Sample project
Included
Best for
Both, kept in sync
Current status
Pre-release

PRE-RELEASE — NOT FOR SALE. No prices are shown — the bundle is the recommended complete system, not a paid upsell here.

The difference

This is not a prompt pack.

Prompts solve single interactions. An operating system holds everything around them.

  • It keeps context and decisions.
  • It distributes work.
  • It defines boundaries.
  • It verifies results.
  • It continues across sessions and models.

Proof

Built from practice, not from theory.

The structure comes from running a real production project — including the parts that break.

  • A real production project, not a theoretical demo.
  • Real deployments to production.
  • Real regressions found and fixed.
  • Real AI handoffs across sessions and models.
  • Real model and resource decisions.
  • Real QA gates before shipping.

Built from the operating system used to develop, audit and run a real production project. No fabricated users, revenue, logos or testimonials.

Buyer questions

What buyers want to know before deciding.

Honest answers to the questions we would hear most often — before there is anything to buy.

Does this work with Claude, ChatGPT, Codex and other tools?

Yes. The kits are provider-neutral: roles, routing and rules are written as Markdown files that you use with any AI model, assistant or editor. Nothing is locked to a specific provider, and no particular model is required.

Do I need programming skills?

No. The kits are structured Markdown files that you read, fill in and add to your project. Git helps but is optional. A text editor is enough.

Are the product files in German or English?

The website is available in German and English — the actual product files and templates are in English. That keeps them usable with every AI tool and in any team.

Is this a SaaS or a download?

A download: a ZIP archive of Markdown files that you use locally inside your project. There is no running online service, nothing is hosted, nothing phones home.

Do I need an account or a cloud?

No. No account, no login, no cloud. The files belong in your project or repository — that is all.

Is this a subscription?

No. A one-time purchase, no recurring usage fees, no consumption-based credits.

What is the difference between the Founder Kit and the Orchestrator Kit?

The Founder Kit organizes the project: context, decisions, status, roadmap, tasks and the next best action. The Orchestrator Kit coordinates the AI team: roles, model routing, working modes, repo gates, reviews and handoffs. One layer runs the project, the other runs the AI tools.

What does the bundle add?

The AI Founder Operating System contains both kits plus the layer that keeps them in sync: the shared file map (SHARED_FILE_MAP), the end-to-end integration flow (INTEGRATION_FLOW), the upgrade path and a fully worked sample project.

Can I use the system in commercial projects?

Yes. The license allows use in an unlimited number of your own private and commercial projects, including adapting and restructuring the files for those projects.

Can I hand the files to clients?

Work you produce with the kit may be used for clients. The original kit files themselves may not be passed on, resold or distributed as a white-label product — including as a client deliverable.

Which updates and which support are included?

All v1.x updates are included for v1 buyers; v2.0 and later major versions are separate products. Support is best-effort email support for installation and reproducible product defects — no consulting, no individual project advice, no customization, no guaranteed response time.

What happens after the purchase?

The later flow: confirm the payment, open the confirmation page, receive the download link, additionally get a delivery email, unzip the archive locally and start with START_HERE_FOR_HUMAN.md (or START_HERE_FOR_AI.md for your agent).

How long does getting started take?

A typical first complete working cycle — set up the files, fill the context, run one task through the loop — takes roughly 25–35 minutes. It depends on your project and the AI tool you use, and it is an orientation, not a guaranteed result.

The products are in pre-release and not yet for sale. Purchase and pricing information follows at launch.

Product evolution

Shipped in versions, not promises.

v1.0

Foundation

The two kits and their shared installation and handoff workflow.

v1.5

Growth

SEO, quality and funnel workflows that measure and improve the work.

v2.0

Cockpit

A neutral, white-label founder cockpit tying everything together — a separate future product under its own terms, not included in a v1 purchase.

See the full roadmap →

Get early access

Products are released in versions. Early adopters get the Foundation versions first. No finished SaaS is promised, and there is no countdown or artificial scarcity.

Request early access