Reset Prototype

Make the engine understandable before we make it clever.

This prototype turns the current Supabase idea into a visible product loop: onboarding, curriculum generation, lesson display, and sandbox practice.

Current reality

  • Supabase has seeded templates and capabilities.
  • An edge function can assemble a first-pass curriculum.
  • The learner experience and data contract still need shape.

Input

Learner snapshot

What the real backend should do

One contract, not scattered guesses

  1. Accept learner profile.
  2. Select level and modules.
  3. Return rendered lesson content plus starter prompts.
  4. Persist progress and sandbox runs later.

Engine view

How the system should fit together

1. Onboarding

Collect role, workflow, readiness, mindset, and maybe constraints.

2. Curriculum assembly

Use `module_templates` and `core_capabilities` to build a learner-specific plan.

3. Guided workspace

Show lessons, starter prompts, success criteria, and module progression.

4. Practice + feedback

Let the learner run prompts, capture output, and score progress over time.

Known issues in the current draft

  • `ai_readiness` mapping likely skips or compresses levels incorrectly.
  • Some placeholders like `{{industry}}` are not supplied by the function.
  • `core_capabilities` is seeded but not yet surfaced in the response.
  • The sandbox is not yet connected to any real tutoring or model runtime.

Lesson

Generate a plan to begin

No module selected
Use the learner snapshot on the left to generate a clearer prototype curriculum.

Sandbox

Practice loop

This is still a UI prototype, not a real model call.
The selected module's starter prompt will appear here.