Build a learning pack
your agent can use.

Your choices shape the path. Free-text use cases and tools stay out of shareable URLs. Never enter credentials or private records.

Start with a pathSafe start Teach one agent Responsible deployment

1. Set the learning contract

Goal:

Experience level:

Subjects:

AI capabilities and limits

Clear prompts and task briefs

Verification and source reading

Everyday privacy and data minimization

Human decisions and approval

AI for learning and studying

Multimodal literacy: images, audio, and text

Teach one agent a bounded role

Knowledge and provenance

Examples, counterexamples, and corrections

Visible memory and forgetting

Structured rules, YAML, and AGENTS.md

Version history and rollback

Creative collaboration

Workplace task design

Least privilege and prompt injection

Evaluations and regression tests

Context, retrieval, visible memory, and model-weight fine-tuning

Responsible deployment and accountability

Portable and self-hosted AI workflows

Use case:Optional and never included in the shareable link. Do not enter private or secret material.

Provider:

Hosting:

Privacy:

Autonomy:

Tools:Optional generic tool names only; never paste keys, tokens or account details.

WHAT YOU RECEIVE

One coherent instruction pack

Prerequisites, boundaries, implementation steps, primary sources and observable tests—ready to copy or download as Markdown.

No hidden personalization.
Only the choices shown in this form affect the result. Nothing is sent to a model.

The agent is the student

Teach one agent.
Inspect everything.

Explore the demo agents →

Teach the Company turns knowledge, examples, corrections and tests into versioned agent files—with human approval before anything leaves the classroom.

The public agents and their histories are fabricated demonstrations. Private installations can connect an OpenAI-compatible or local model; the website does not claim model-weight fine-tuning.

Demo agents AI school Public manuals Microcontroller projects Suggest a correction Docker self-hosting Privacy & rules Open source Changelog

Build a learning packyour agent can use.

1. Set the learning contract

Build a learning pack
your agent can use.