implement

Pattern-Guided Implementation Workflow

You are implementing the feature: $ARGUMENTS

Intelligence Layer (ACTIVATE FIRST)

1. Load PRD context: specs/active/{slug}/prd.md
2. Load patterns: specs/active/{slug}/patterns/ and specs/guides/patterns/
3. Load task list: specs/active/{slug}/tasks.md
4. Check MCP strategy: .claude/mcp-strategy.md

---

Checkpoint 0: Context Loading

Required reads:

cat specs/active/{slug}/prd.md
cat specs/active/{slug}/tasks.md
cat specs/active/{slug}/research/analysis.md

Verify PRD exists and is complete.

Output: "✓ Checkpoint 0 complete - PRD loaded, [N] tasks identified"

---

Checkpoint 1: Pattern Deep Dive

Read 3-5 similar implementations identified in PRD:

For Litestar backend:

• Read existing controller in same domain
• Read existing service in same domain
• Read related models

For React/Inertia frontend:

• Read similar page component
• Read related UI components

Document patterns in specs/active/{slug}/tmp/patterns-found.md:

## Patterns to Follow

### Controller Pattern
- Uses `@Controller` decorator with path
- Injects services via constructor
- Returns `InertiaResponse` for pages

### Service Pattern
- Extends `SQLAlchemyAsyncRepositoryService[Model]`
- Has `repository_type` class attribute
- Custom methods for business logic

### Repository Pattern
- Extends from advanced-alchemy base
- Uses model type parameter

### Frontend Pattern
- Page receives typed props
- Uses shadcn/ui components
- Follows resources/pages/{feature}/ structure

Output: "✓ Checkpoint 1 complete - Patterns documented"

---

Checkpoint 2: Database Layer (if needed)

Create models in app/db/models/:

Follow existing model patterns:

• Use __tablename__ with snake_case
• Include __table_args__ if needed
• Use proper type hints with SQLAlchemy 2.0 syntax
• Export in app/db/models/__init__.py

Create migration:

uv run app database make-migrations