task-breakdown

Task Breakdown

Feature: $0

Granularity: $1 (default: detailed)

Steps

1. Understand the scope — Read user stories, PRDs, or feature descriptions. If code exists, read the relevant areas to understand current state and what needs to change.

2. Identify work streams:

   • Backend: API endpoints, database changes, business logic
   • Frontend: components, pages, forms, state management
   • Infrastructure: deploy config, environment variables, CI changes
   • Testing: unit, integration, e2e tests
   • Documentation: README updates, API docs, ADRs

3. Break into tasks — Each task should be:

   • Completable by one person in 1-3 days
   • Independently mergeable (or clearly part of a sequence)
   • Testable on its own

4. Map dependencies — Which tasks block others?

   • Database schema must exist before API endpoints
   • API endpoints must exist before frontend integration
   • But: frontend components can be built with mocks in parallel

5. Suggest implementation order — Optimize for:

   • Unblocking parallel work early
   • Getting a vertical slice working end-to-end first
   • Reducing risk (hard/uncertain tasks first)

Output Format

## Task Breakdown: <feature>

### Summary
- **Total tasks**: N
- **Estimated effort**: X days (1 developer)
- **Parallelizable**: Y (with 2 developers, ~Z days)

### Critical Path
`[Schema] → [API] → [Frontend Integration] → [E2E Tests]`

### Tasks

#### Phase 1: Foundation (unblocks everything)

- [ ] **T-001**: <task title>
  - **Stream**: backend | frontend | infra | testing
  - **Estimate**: S (< 1 day) | M (1-2 days) | L (2-3 days) | XL (3+ days, should be split)
  - **Depends on**: — (none)
  - **Details**: <what specifically needs to happen>

- [ ] **T-002**: <task title>
  - **Stream**: backend
  - **Estimate**: M
  - **Depends on**: T-001
  - **Details**: <specifics>

#### Phase 2: Core Implementation (can parallelize)
...

#### Phase 3: Polish & Testing
...

### Parallel Work Opportunities
- Developer A: T-001 → T-003 → T-005
- Developer B: T-002 → T-004 → T-006

### Risks
- <risk>: <mitigation>

Constraints

• Tasks should be small enough to review in one PR
• XL tasks (3+ days) should be split further
• Always include testing tasks — they're not optional
• Don't create tasks for trivial work (e.g., "create file" — that's part of the implementation task)
• Include infra/config tasks that are easy to forget (env vars, migrations, CI updates)