Files
gh-ninthspace-claude-code-m…/commands/story-new.md
2025-11-30 08:44:27 +08:00

359 lines
11 KiB
Markdown

# /sdd:story-new
## Meta
- Version: 2.0
- Category: story-management
- Complexity: medium
- Purpose: Create new story with auto-populated template and place in backlog
## Definition
**Purpose**: Create a new story using project context and place it in the backlog folder for future development.
**Syntax**: `/sdd:story-new [story_id_number]`
## Parameters
| Parameter | Type | Required | Default | Description | Validation |
|-----------|------|----------|---------|-------------|------------|
| story_id_number | number | No | auto-increment | Story number (used as STORY-YYYY-NNN) | Positive integer |
## INSTRUCTION: Create New Story
### INPUTS
- story_id_number: Optional story number (auto-increments if not provided)
- Project context from `/docs/project-context/` directory
- User-provided story details (if not in project brief)
### PROCESS
#### Phase 1: Project Context Loading
1. **CHECK** if `/docs/project-context/` directory exists
2. IF missing:
- SUGGEST running `/sdd:project-init` first
- EXIT with initialization guidance
3. **LOAD** project context from:
- `/docs/project-context/project-brief.md` - Existing story definitions, goals
- `/docs/project-context/technical-stack.md` - Technology implementation requirements
- `/docs/project-context/coding-standards.md` - Testing and quality requirements
#### Phase 2: Story ID Generation
1. **GENERATE** story ID in format `STORY-YYYY-NNN`:
- YYYY = current year
- NNN = sequential number (001, 002, etc.)
2. IF user provides story_id_number:
- USE as basis: `STORY-YYYY-[story_id_number]`
- EXAMPLE: Input "5" → "STORY-2025-005"
3. **CHECK** for existing IDs across all directories:
- SCAN `/docs/stories/backlog/`
- SCAN `/docs/stories/development/`
- SCAN `/docs/stories/review/`
- SCAN `/docs/stories/qa/`
- SCAN `/docs/stories/completed/`
- CHECK `/docs/project-context/project-brief.md` for planned stories
4. IF no specific number provided:
- INCREMENT to next available number
- ENSURE uniqueness across all locations
#### Phase 3: Story Information Gathering
1. **SEARCH** `/docs/project-context/project-brief.md` for story with generated ID
2. IF story exists in project brief:
- EXTRACT comprehensive story details:
* Story title and description
* User scenarios and use cases
* Technical implementation requirements
* Acceptance criteria (pass/fail conditions)
* Edge cases and error handling requirements
* UI/UX considerations
* Testing requirements and test scenarios
* Integration points with other stories/systems
* Dependencies on other stories
3. IF story NOT found in project brief:
- **ASK** user for story title
- **ASK** user for story description and purpose
- **ASK** user for acceptance criteria
- **ASK** user for technical approach (optional)
- **ASK** user for dependencies (optional)
#### Phase 4: Story File Creation
1. **ENSURE** `/docs/stories/backlog/` directory exists
- CREATE directory if missing
- ADD `.gitkeep` file if directory was created
2. **CREATE** story file at `/docs/stories/backlog/[story-id].md`
3. **POPULATE** template with:
- Story ID and title
- Status: backlog
- Today's date as "Started" date
- Empty "Completed" date
- Note: "(none - in backlog)" for branch
- What & Why section (from project brief or user input)
- Success Criteria (from project brief or user input)
- Technical Notes with:
* Approach (from project brief or user input)
* Stack (auto-populated from technical-stack.md)
* Concerns (from project brief edge cases or user input)
* Dependencies (from project brief or user input)
- Implementation Checklist (standard items)
- Progress Log with creation entry
- Test Cases (from project brief or default scenarios)
- UI/UX Considerations (from project brief if applicable)
- Integration Points (from project brief if applicable)
- Rollback Plan section (empty template)
- Lessons Learned section (empty template)
4. **REFERENCE** project context in template:
- Pull technology stack from technical-stack.md
- Note coding standards that will apply
- Reference testing framework requirements
- Include project goals and constraints from project-brief.md
#### Phase 5: Completion Summary
1. **DISPLAY** creation summary:
```
✅ Story Created
═══════════════════════════════════
Story ID: [STORY-YYYY-NNN]
Title: [Story Title]
Location: /docs/stories/backlog/[story-id].md
Status: backlog
[If from project brief:]
Source: Extracted from project brief
- Acceptance criteria: [count] criteria defined
- Test scenarios: [count] scenarios defined
- Dependencies: [list or "None"]
[If from user input:]
Source: User-provided details
- Ready for refinement before development
```
2. **SUGGEST** next steps:
```
💡 NEXT STEPS:
1. /sdd:story-start [story-id] # Move to development and create branch
2. /sdd:story-implement [story-id] # Generate implementation code
3. /sdd:project-status # View all project stories
```
### OUTPUTS
- `/docs/stories/backlog/[story-id].md` - New story file with populated template
- `.gitkeep` file in `/docs/stories/backlog/` if directory was created
### RULES
- MUST generate unique story ID across all story directories
- MUST create backlog directory if it doesn't exist
- MUST auto-populate template with project context
- SHOULD extract story details from project brief if available
- SHOULD reference technical stack in story template
- NEVER create feature branch (stories start in backlog)
- ALWAYS add progress log entry for creation
- MUST include today's date as "Started" date
## Story Template Structure
```markdown
# [STORY-ID]: [Title]
## Status: backlog
**Started:** [Today's Date]
**Completed:**
**Branch:** (none - in backlog)
## What & Why
[Story description and purpose]
## Success Criteria
- [ ] [Criterion 1]
- [ ] [Criterion 2]
- [ ] [Criterion 3]
## Technical Notes
**Approach:** [Implementation approach]
**Stack:** [Auto-populated from technical-stack.md]
**Concerns:** [Risks and edge cases]
**Dependencies:** [External services/libraries/other stories]
## Implementation Checklist
- [ ] Feature implementation
- [ ] Unit tests
- [ ] Integration tests
- [ ] Error handling
- [ ] Loading states
- [ ] Documentation
- [ ] Performance optimization
- [ ] Accessibility
- [ ] Security review
## Progress Log
- [Today]: Created story, added to backlog
## Test Cases
1. Happy path: [scenario]
2. Error case: [scenario]
3. Edge case: [scenario]
## UI/UX Considerations
[User interface and experience requirements]
## Integration Points
[Dependencies and integration with other systems]
## Rollback Plan
[How to rollback if issues arise]
## Lessons Learned
[To be filled when complete]
```
## Examples
### Example 1: Create from Project Brief
```bash
INPUT:
/sdd:story-new
OUTPUT:
→ Checking project context...
→ Generating story ID: STORY-2025-001
→ Found story definition in project brief
✅ Story Created
═══════════════════════════════════
Story ID: STORY-2025-001
Title: User Authentication System
Location: /docs/stories/backlog/STORY-2025-001.md
Status: backlog
Source: Extracted from project brief
- Acceptance criteria: 5 criteria defined
- Test scenarios: 8 scenarios defined
- Dependencies: None
💡 NEXT STEPS:
1. /sdd:story-start STORY-2025-001 # Move to development and create branch
2. /sdd:story-implement STORY-2025-001 # Generate implementation code
3. /sdd:project-status # View all project stories
```
### Example 2: Create with Specific ID
```bash
INPUT:
/sdd:story-new 10
OUTPUT:
→ Checking project context...
→ Using story ID: STORY-2025-010
→ Story not found in project brief, gathering details...
What is the story title?
> Add Dark Mode Toggle
What are you building and why?
> Implement a dark mode toggle in the settings page to allow users to switch between light and dark themes.
What are the acceptance criteria? (Enter each, then empty line when done)
> Toggle is visible in settings page
> Theme persists across sessions
> All UI components support both themes
>
✅ Story Created
═══════════════════════════════════
Story ID: STORY-2025-010
Title: Add Dark Mode Toggle
Location: /docs/stories/backlog/STORY-2025-010.md
Status: backlog
Source: User-provided details
- Ready for refinement before development
💡 NEXT STEPS:
1. /sdd:story-start STORY-2025-010 # Move to development and create branch
2. /sdd:story-implement STORY-2025-010 # Generate implementation code
3. /sdd:project-status # View all project stories
```
### Example 3: Auto-Increment ID
```bash
INPUT:
/sdd:story-new
OUTPUT:
→ Checking project context...
→ Found existing stories: STORY-2025-001 through STORY-2025-005
→ Auto-incrementing to: STORY-2025-006
→ Story not found in project brief, gathering details...
[Interactive prompts for story details...]
✅ Story Created
═══════════════════════════════════
Story ID: STORY-2025-006
Title: Payment Processing Integration
Location: /docs/stories/backlog/STORY-2025-006.md
Status: backlog
```
## Edge Cases
### No Project Context
- DETECT missing `/docs/project-context/` directory
- SUGGEST running `/sdd:project-init`
- OFFER to create story with minimal template
- WARN that template won't be auto-populated
### Duplicate Story ID
- DETECT ID conflict across all directories
- INCREMENT to next available number automatically
- LOG warning about skipped number
- ENSURE final ID is unique
### Empty Project Brief
- DETECT missing story definitions
- GATHER all details from user interactively
- CREATE story with user-provided information
- SUGGEST adding stories to project brief
### Malformed Project Brief
- DETECT parsing errors
- LOG warning about brief issues
- FALL BACK to user input mode
- CONTINUE with story creation
## Error Handling
- **Missing /docs/project-context/**: Suggest `/sdd:project-init` with guidance
- **Permission errors**: Report specific file/directory with access issue
- **Invalid story ID**: Sanitize and suggest corrected version
- **User cancels**: Clean up partial creation, exit gracefully
## Performance Considerations
- Story ID checking optimizes by scanning directories once
- Project brief parsing caches results for session
- Template population is fast (< 100ms typically)
- Interactive prompts allow user to control pace
## Related Commands
- `/sdd:project-init` - Initialize project structure first
- `/sdd:project-brief` - Create/update project documentation with stories
- `/sdd:story-start [id]` - Begin development on story
- `/sdd:story-implement [id]` - Generate implementation code
- `/sdd:project-status` - View all project stories
## Constraints
- ✅ MUST generate unique story ID
- ✅ MUST create story in backlog directory
- ⚠️ NEVER create feature branch (stories start in backlog)
- 📋 MUST auto-populate from project context when available
- 🔧 SHOULD extract from project brief before asking user
- 💾 MUST add creation entry to progress log
- 📅 MUST include today's date as "Started" date