SystangoTechnologies/planning-and-task-breakdown

Planning and Task Breakdown

Overview

Decompose work into small, verifiable tasks with explicit acceptance criteria. Good task breakdown is the difference between an agent that completes work reliably and one that produces a tangled mess. Every task should be small enough to implement, test, and verify in a single focused session.

When to Use

• You have a spec and need to break it into implementable units
• A task feels too large or vague to start
• Work needs to be parallelized across multiple agents or sessions
• You need to communicate scope to a human
• The implementation order isn't obvious

When NOT to use: Single-file changes with obvious scope, or when the spec already contains well-defined tasks.

The Planning Process

Step 1: Enter Plan Mode

Before writing any code, operate in read-only mode:

• Read .context/project.md and .context/architecture.md for system context
• Read .context/conventions.md to identify existing patterns and standards to follow
• Read .context/concerns.md to surface risks and unknowns before the plan is written
• Read the spec and relevant codebase sections
• Map dependencies between components

Do NOT write code during planning. The output is a plan document, not implementation.

Step 2: Identify the Dependency Graph

Map what depends on what:

Database schema
    │
    ├── API models/types
    │       │
    │       ├── API endpoints
    │       │       │
    │       │       └── Frontend API client
    │       │               │
    │       │               └── UI components
    │       │
    │       └── Validation logic
    │
    └── Seed data / migrations

Implementation order follows the dependency graph bottom-up: build foundations first.

Step 3: Slice Vertically

Instead of building all the database, then all the API, then all the UI — build one complete feature path at a time:

Bad (horizontal slicing):

Task 1: Build entire database schema
Task 2: Build all API endpoints
Task 3: Build all UI components
Task 4: Connect everything

Good (vertical slicing):

Task 1: User can create an account (schema + API + UI for registration)
Task 2: User can log in (auth schema + API + UI for login)
Task 3: User can create a task (task schema + API + UI for creation)
Task 4: User can view task list (query + API + UI for list view)

Each vertical slice delivers working, testable functionality.

Step 4: Write Tasks

Planning produces two documents, not one:

• specs/tasks/plan.md — the high-level plan. Contains the overview, architecture decisions, a compact Task Index (one line per task), checkpoints, risks, and open questions. See the Plan Document Template below.
• specs/tasks/todo.md — the detailed task list. Contains the full per-task block (template below) for every task in the index, stacked in order. See the todo.md Template below.

Think of plan.md as the table of contents and todo.md as the chapters. Every task in plan.md's Task Index must have a matching detailed entry in todo.md.

Each task in todo.md follows this structure:

## Task [N]: [Short descriptive title]

**Description:** One paragraph explaining what this task accomplishes.

**Acceptance criteria:**
- [ ] [Specific, testable condition]
- [ ] [Specific, testable condition]

**Unit Tests (deferred):**
- [ ] Tests written
- Functions/behaviors to cover: [e.g. createTask, validateInput]
- Edge cases: [e.g. empty title, duplicate ID, missing required field]
- Test type: [unit | component | end-to-end]
- Suggested file: `tests/path/to/feature.test.ts`

**Verification:**
- [ ] Build succeeds: `npm run build`
- [ ] Type checking passes: `npx tsc --noEmit`
- [ ] Manual check: [description of what to verify]

**Dependencies:** [Task numbers this depends on, or "None"]

**Files likely touched:**
- `src/path/to/file.ts`

**Estimated scope:** [Small: 1-2 files | Medium: 3-5 files | Large: 5+ files]

**Domain skill:** [`api-and-interface-design` | `frontend-ui-engineering` | `security-and-hardening` | None]

Field-by-field rules:

• Acceptance criteria — every item must use checkbox syntax (- [ ]), not plain bullets (- ). Plain bullets are not trackable.
• Unit Tests (deferred) — required for every task. The - [ ] Tests written line must use checkbox syntax (- [ ]), not a plain bullet. This checkbox is the contract between the planning and testing phases — the /test phase checks it after writing tests. The build phase skips this section entirely and does not create test files.
• Verification — every item must use checkbox syntax (- [ ]), not plain bullets. Build-phase gates verify compilation and runtime only — never npm test, pnpm test, or coverage thresholds.
• Domain skill — required for every task. The planner knows the task's domain when writing it. Embed the classification so the builder doesn't have to re-derive it. Use None for pure infrastructure tasks (build config, utility functions, migrations with no auth surface). Tasks involving REST endpoints, HTTP status codes, request/response contracts, or middleware must specify api-and-interface-design. Tasks involving UI components, layouts, or state must specify frontend-ui-engineering. Tasks involving auth, validation, PII, or sessions must specify security-and-hardening. A task may list more than one skill if it spans domains.

todo.md Template

todo.md is a flat stack of detailed task blocks — no extra prose, no simplified bullets. Every task from plan.md's Task Index appears here in full form, in the same order:

# Task List: [Feature/Project Name]

Detailed tasks for the plan in `plan.md`. Build one task at a time, top to bottom.

---

## Task 1: [Short descriptive title]

**Description:** One paragraph explaining what this task accomplishes.

**Acceptance criteria:**
- [ ] [Specific, testable condition]
- [ ] [Specific, testable condition]

**Unit Tests (deferred):**
- [ ] Tests written
- Functions/behaviors to cover: [e.g. createTask, validateInput]
- Edge cases: [e.g. empty title, duplicate ID, missing required field]
- Test type: [unit | component | end-to-end]
- Suggested file: `tests/path/to/feature.test.ts`

**Verification:**
- [ ] Build succeeds: `npm run build`
- [ ] Type checking passes: `npx tsc --noEmit`
- [ ] Manual check: [description of what to verify]

**Dependencies:** None

**Files likely touched:**
- `src/path/to/file.ts`

**Estimated scope:** Small: 1-2 files

**Domain skill:** None

---

## Task 2: [Short descriptive title]

**Description:** ...

**Acceptance criteria:**
- [ ] ...

**Unit Tests (deferred):**
- [ ] Tests written
- Functions/behaviors to cover: ...
- Edge cases: ...
- Test type: unit
- Suggested file: `tests/path/to/feature.test.ts`

**Verification:**
- [ ] Build succeeds: `npm run build`
- [ ] Type checking passes: `npx tsc --noEmit`
- [ ] Manual check: ...

**Dependencies:** Task 1

**Files likely touched:**
- `src/path/to/file.ts`

**Estimated scope:** Medium: 3-5 files

**Domain skill:** api-and-interface-design

---

## Checkpoint: After Tasks 1-2
- [ ] Build passes
- [ ] Core flow works end-to-end

---

## Task 3: ...

Every task block must include all fields from the per-task template above. Do not abbreviate, do not use simplified bullets, do not omit the Unit Tests (deferred) section. If you find yourself shortening a task block to save space, stop — either the task belongs in the Task Index only (plan.md) or it needs its full detail here.

Step 5: Validate Task List Structure

Before advancing to the build phase, run this checklist against every task in the list. If any item fails, fix the task before proceeding — do not defer fixes to the build phase.

Phase separation:

• [ ] Every task has a Unit Tests (deferred) section with - [ ] Tests written using checkbox syntax (- [ ]), not a plain bullet (- )
• [ ] No task's primary deliverable is a test file — test files are created in the /test phase, not as build tasks. This includes tasks titled "Test Suite", "Write Tests", etc. — if its purpose is producing test files, it should not exist as a task
• [ ] No task description mentions creating test files, placeholder tests, or dummy test files during the build phase
• [ ] No gate or verification step requires npm test, pnpm test, --listTests, or coverage thresholds — build gates verify compilation and runtime only

Template compliance:

• [ ] Every acceptance criterion uses checkbox format (- [ ]), not plain bullets (- )
• [ ] Every verification step uses checkbox format (- [ ]), not plain bullets (- )
• [ ] Every task that involves API, UI, or security work has a **Domain skill:** field identifying which skill to load before implementation. Tasks with REST endpoints, HTTP status codes, request/response contracts, or middleware must specify api-and-interface-design. Pure infrastructure tasks use None

File location:

• [ ] Task list is saved to specs/tasks/todo.md

Step 6: Order and Checkpoint

Arrange tasks so that:

1. Dependencies are satisfied (build foundation first)
2. Each task leaves the system in a working state
3. Verification checkpoints occur after every 2-3 tasks
4. High-risk tasks are early (fail fast)

Add explicit checkpoints:

## Checkpoint: After Tasks 1-3
- [ ] All tests pass
- [ ] Application builds without errors
- [ ] Core user flow works end-to-end
- [ ] Review with human before proceeding

Task Sizing Guidelines

| Size | Files | Scope | Example | |------|-------|-------|---------| | XS | 1 | Single function or config change | Add a validation rule | | S | 1-2 | One component or endpoint | Add a new API endpoint | | M | 3-5 | One feature slice | User registration flow | | L | 5-8 | Multi-component feature | Search with filtering and pagination | | XL | 8+ | Too large — break it down further | — |

If a task is L or larger, it should be broken into smaller tasks. An agent performs best on S and M tasks.

When to break a task down further:

• It would take more than one focused session (roughly 2+ hours of agent work)
• You cannot describe the acceptance criteria in 3 or fewer bullet points
• It touches two or more independent subsystems (e.g., auth and billing)
• You find yourself writing "and" in the task title (a sign it is two tasks)

Plan Document Template

This is the template for plan.md only. The Task Index below is a compact reference — one line per task. The full detailed task blocks go in todo.md (see the todo.md Template in Step 4).

# Implementation Plan: [Feature/Project Name]

## Overview
[One paragraph summary of what we're building]

## Architecture Decisions
- [Key decision 1 and rationale]
- [Key decision 2 and rationale]

## Task Index

Compact index of tasks. Full detail for each task lives in `todo.md`.

### Phase 1: Foundation
- [ ] Task 1: ...
- [ ] Task 2: ...

### Checkpoint: Foundation
- [ ] Tests pass, builds clean

### Phase 2: Core Features
- [ ] Task 3: ...
- [ ] Task 4: ...

### Checkpoint: Core Features
- [ ] End-to-end flow works

### Phase 3: Polish
- [ ] Task 5: ...
- [ ] Task 6: ...

### Checkpoint: Complete
- [ ] All acceptance criteria met
- [ ] Ready for review

## Risks and Mitigations
| Risk | Impact | Mitigation |
|------|--------|------------|
| [Risk] | [High/Med/Low] | [Strategy] |

## Open Questions
- [Question needing human input]

Parallelization Opportunities

When multiple agents or sessions are available:

• Safe to parallelize: Independent feature slices, tests for already-implemented features, documentation
• Must be sequential: Database migrations, shared state changes, dependency chains
• Needs coordination: Features that share an API contract (define the contract first, then parallelize)

Common Rationalizations

| Rationalization | Reality | |---|---| | "I'll figure it out as I go" | That's how you end up with a tangled mess and rework. 10 minutes of planning saves hours. | | "The tasks are obvious" | Write them down anyway. Explicit tasks surface hidden dependencies and forgotten edge cases. | | "Planning is overhead" | Planning is the task. Implementation without a plan is just typing. | | "I can hold it all in my head" | Context windows are finite. Written plans survive session boundaries and compaction. | | "I'll make the tests a separate task" | Tests belong in the /test phase. Each task's Unit Tests (deferred) section captures what to test. Creating a test task defeats phase separation and forces tests to be written during build. Don't do this. |

Red Flags

• Starting implementation without a written task list
• Tasks that say "implement the feature" without acceptance criteria
• No verification steps in the plan
• Unit Tests (deferred) section missing from a task (every task needs test notes, even if deferred)
• Tests written line using a plain bullet (- Tests written) instead of checkbox syntax (- [ ] Tests written) — the checkbox is the contract between planning and testing phases
• Acceptance criteria or verification steps using plain bullets instead of checkbox syntax — plain bullets are not trackable
• Task whose primary deliverable is a test file — tests are a phase, not a task. Document them in Unit Tests (deferred), don't build them as tasks
• Task description mentioning "create test files", "placeholder tests", or "dummy test files" — no test files are created during build
• Gate or verification step that requires npm test, pnpm test, --listTests, or coverage thresholds — build gates verify compilation and runtime, not test coverage
• API/UI/security task missing a Domain skill field — the planner knows the domain; don't make the builder guess
• All tasks are XL-sized
• No checkpoints between tasks
• Dependency order isn't considered

Verification

Before starting implementation, confirm:

• [ ] Every task has acceptance criteria using checkbox syntax (- [ ])
• [ ] Every task has verification steps using checkbox syntax (- [ ])
• [ ] Every task has a Unit Tests (deferred) section with - [ ] Tests written (checkbox, not plain bullet) and at least one behavior listed
• [ ] Every task has a **Domain skill:** field (api-and-interface-design, frontend-ui-engineering, security-and-hardening, or None)
• [ ] No task's primary deliverable is a test file
• [ ] No gate requires npm test, pnpm test, or coverage thresholds
• [ ] Task dependencies are identified and ordered correctly
• [ ] No task touches more than ~5 files
• [ ] Checkpoints exist between major phases
• [ ] The human has reviewed and approved the plan