Pipeline-driven interactive fiction generation.
Pre-implementation — Design documentation complete, implementation pending.
See docs/design/ for the complete v5 architecture specification.
| Document | Description |
|---|---|
| 00-vision.md | Core philosophy and pipeline overview |
| 01-pipeline-architecture.md | Orchestrator, stages, gates |
| 02-artifact-schemas.md | YAML artifact formats |
| 03-grow-stage-specification.md | GROW stage six-layer decomposition |
| 04-state-mechanics.md | Codewords, stats, state management |
| 05-prompt-compiler.md | Prompt template system |
| 06-quality-bars.md | Validation criteria |
| 07-design-principles.md | Core design principles |
| 08-research-foundation.md | Research basis |
| 09-v4-reference.md | Historical context (non-canonical) |
| 10-semantic-conventions.md | Naming conventions |
| 11-proposed-dependencies.md | Proposed tech stack |
| 12-getting-started.md | Implementation guide |
| 13-project-structure.md | Directory layout |
DREAM → BRAINSTORM → SEED → GROW → FILL → SHIP
QuestFoundry uses an ontology-first approach: the graph ontology defines node types and relationships, and hand-written Pydantic models in src/questfoundry/models/ implement it. The graph (graph.db) is the runtime source of truth.
TBD