# Initial Prototype Roadmap

## Prototype 0.1
Goal: prove the orchestration model.

Capabilities:
- accept domain and goal
- generate a toy concept graph
- produce a staged roadmap
- generate one Socratic prompt
- generate one practice task
- generate one capstone suggestion
- produce one evaluation rubric
- list locally installed domain packs

## Prototype 0.2
Goal: learner-state adaptation.

Capabilities:
- accept prior knowledge signals
- adjust roadmap depth
- store evidence items
- identify weakest competencies
- load a selected domain pack

## Prototype 0.3
Goal: local-model integration.

Capabilities:
- connect to local Ollama endpoint
- select configured model
- use deterministic prompts for component outputs
- retain remote provider fallback but disable by default
- validate artifact compatibility