Generate Implementation Plan

Generate Implementation Plan

Overview Document: $ARGUMENTS

Overview

This workflow creates two outputs:

1. Implementation Plan (Confluence) - Coordination dashboard for tracking execution
2. Updated Jira Tickets - Self-contained with full implementation details

Phase 0: Context Retrieval

1. Fetch the overview document from {Overview_Document}

2. Extract:

   • {Epic_Key} - The epic ticket key (e.g., CC-123)
   • {Jira_Project} - Link to the Jira project/board
   • Related ticket links

3. FAILURE CONDITION: If epic key or project cannot be determined, prompt user for missing info.

4. CHECKPOINT: Confirm epic key and project before proceeding.

Phase 1: Discovery

1. Read the overview document for epic context, scope, and requirements
2. Fetch all Jira tickets linked to this epic
3. Explore the codebase for patterns relevant to the tickets

Codebase Exploration Focus:

• Affected files/modules per ticket
• API patterns (routes, middleware, response formats)
• Component patterns (naming, state management)
• Test patterns (locations, fixtures, mocks)
• Reference implementations

Phase 2: Ticket Refinement

Analyze existing tickets and prepare adjustments:

• Create new tickets for gaps (missing functionality, dependencies)
• Split large tickets (>8 story points)
• Add story points to all tickets
• Link dependencies (blocked-by relationships)

CHECKPOINT: Present proposed ticket changes and get approval before modifying Jira.

Phase 3: Generate Implementation Plan

Create a lightweight coordination document in Confluence.

Implementation Plan Structure

# {Epic_Key} Implementation Plan

## Summary

| Metric | Value |
|--------|-------|
| **Epic** | [{Epic_Key}](link) |
| **Total Tickets** | X |
| **Total Story Points** | X |
| **Overall Complexity** | Low/Medium/High |
| **Execution Waves** | X |
| **Key Dependencies** | Brief summary |

---

## Completion Status

| Ticket | Status | Completed Date | Notes |
|--------|--------|----------------|-------|
| [CC-XXX](link) | Pending | - | - |
| [CC-YYY](link) | Pending | - | - |

---

## Execution Order (Topologically Sorted)

| # | Ticket | Summary | Points | Risk | Dependencies | Status |
|---|--------|---------|--------|------|--------------|--------|
| 1 | [CC-XXX](link) | Title | 2 | Low | None | Pending |
| 2 | [CC-YYY](link) | Title | 3 | Low | CC-XXX | Pending |

---

## Parallel Execution Strategy

### Wave 1: [Name] (X pts) - Execute in Parallel

| Ticket | Summary | Points | Agent | Status |
|--------|---------|--------|-------|--------|
| CC-XXX | Title | 2 | frontend-developer | Pending |
| CC-YYY | Title | 3 | backend-developer | Pending |

### Wave 2: [Name] (X pts) - After Wave 1

| Ticket | Summary | Points | Agent | Status |
|--------|---------|--------|-------|--------|
| CC-ZZZ | Title | 5 | fullstack-developer | Pending |

**Note:** [Any blocking dependencies or coordination notes]

---

## Agent Recommendations

| Work Type | Recommended Agent |
|-----------|-------------------|
| Frontend UI/components | frontend-developer |
| API/backend services | backend-developer |
| Database changes | database-optimizer |
| Full-stack features | fullstack-developer |
| Test coverage | test-automator |

---

## Document Links

- **Overview Document:** [link]
- **Epic:** [link]

CHECKPOINT: Present plan preview and get approval before publishing.

Phase 4: Update Jira Tickets

After publishing the implementation plan, update each ticket with full implementation details.

Ticket Description Template

Each ticket must be self-contained with everything needed to execute:

Overview Document: {Overview_Document_URL}
Implementation Plan: {Implementation_Plan_URL}

## Summary

[Brief description of what this ticket accomplishes]

## Implementation Steps

1. **[Step title]**
   - Specific action with file path
   - Code snippet if applicable
   ```typescript
   // Example code

2. [Step title]
   • Specific action
   • Details

Files to Modify

Tests

File: tests/unit/feature.test.ts

// Complete, copy-paste ready test code
import { ... } from '...';

describe('Feature', () => {
  beforeEach(() => {
    // setup
  });

  test('should do X', () => {
    // test implementation
  });
});

Acceptance Criteria

• Implementation step 1 complete
• Implementation step 2 complete
• All tests passing
• No TypeScript errors (npm run check)
• Linting passes (npm run lint)

### Why Self-Contained Tickets

- `/execute-ticket` can run with just the ticket (no external doc fetching required)
- All context in one place
- Reduces token usage during execution
- Enables parallel agent execution

---

## Phase 5: Update Overview Document

Add an **Adjustments Section** to the overview document:

```markdown
## Implementation Plan Adjustments

**Plan Created:** [date]
**Implementation Plan:** [link]

### Tickets Added
- [CC-XXX] - [title] - [justification]

### Tickets Split
- [CC-YYY] split into [CC-YYY1], [CC-YYY2] - [reason]

### Story Points Updated
- [CC-ZZZ]: X → Y points - [reason]

### New Dependencies
- [CC-AAA] blocked by [CC-BBB] - [reason]

CHECKPOINT: Present proposed updates and get approval before modifying.

Output

When complete, provide:

## Implementation Plan Created

**Implementation Plan:** [Confluence URL]
**Overview Document:** [Updated URL]

### Summary
- Tickets in epic: X
- Total story points: X
- Execution waves: X

### Changes Made
- Tickets created: [list]
- Tickets updated: [list]
- Story points added: [list]

### Next Steps
1. Review the implementation plan
2. Run `/execute-ticket <ticket-key>` to begin work

Checkpoints Summary

Never modify Jira or Confluence without explicit user approval.