Practice with purpose

Learn through katas.

Pick something you want to make. Follow a short recipe and try it with an AI assistant.

Build a prototype Turn an idea into a plan

Start now

Your first two katas

Choose the result you want. Both are beginner friendly.

About 15 minutes

Make a clickable prototype

Build a small interactive experience and make it your own.

Start the prototype kata

About 10 minutes

Turn an idea into a plan

Use an AI assistant to turn a spark into a clear one-page project brief.

Start the idea kata

One foundation, many fields

Learn the stack. Apply it anywhere.

The orchestrator is the developer. Shared foundations include harnesses, models, instructions, context, memory, skills, integrations, delegation and verification. You choose and equip the system, then take responsibility for the outcome.

Specializations add the artifacts, tools, standards and review needed in each field. The same foundation can support software, UX, 3D and BIM, construction, science, manufacturing, education and more. The catalog is being developed; these fields are directions for future practice, not a claim that all are available today.

In safety-critical or regulated work, an agent-generated artifact needs independent validation and any required approval from qualified professionals. A Dojo belt does not grant a professional license.

Your route through the Dojo

Choose where to begin

Explore the spaces that support practice. Collections, demonstrations and assessments will grow as their evidence is verified.

See an outcome

Start with real examples as the Gallery grows.

Explore the Gallery

Learn the stack

Study the foundation shared by every field.

Enter the Stack Lab

Choose a field

See what changes when a technique enters a new domain.

Explore fields

Equip your agents

Learn how harnesses, skills and integrations fit together.

Visit the Armory

Compare with evidence

Understand how to evaluate a complete stack.

Visit the Arena

Demonstrate capability

Discover the vision for domain-specific belts.

Explore belts

Try an unfamiliar challenge

Choose your own approach through Randori.

Explore Randori

Apply and share

Take practice into real work and learn from the outcome.

Explore missions

The practice

From outcome to independent mastery

The kata experience is being built. This is the learning journey it is designed to support.

See an outcome and study the kata. Start with a real result, then work backward through its source, stack and evidence.
Assemble the stack and equip the agents. Choose a harness, model, instructions, context, skills and permitted tools for the task.
Execute and verify. Follow the recipe, check the outcome against clear criteria and investigate failures.
Adapt. Change a constraint, tool or goal and learn which parts of the technique transfer.
Demonstrate capability. Future domain belts will assess evidence of skill; Randori will challenge you to choose your own approach to an unfamiliar goal.
Apply and teach. Missions put techniques to work; Senseis help others improve and refine the katas.

First foundation kata in development

Same goal, different stack

The first planned kata will compare how a chosen outcome changes as a learner varies a harness, model, instruction or other stack component. It will use a real source exercise and recorded evidence, with controlled comparisons wherever practical. The exercise and its verified recipe are still being prepared.

While it is in development, explore the existing Working with AI agents track and free resource library. The existing Developer and Master Developer training remains available.

Still discovering what interests you? Explore with Ignite Curiosity, our sibling learning community. You can start at either site.