task-details（任务详情）

Task Details Enrichment Skill

Automatically enrich Jira tasks with comprehensive context, technical requirements, dependencies, and estimates.

When to Use

• Issue lacks sufficient detail for implementation
• Need to extract hidden requirements from descriptions
• Analyzing issue complexity for estimation
• Identifying dependencies and blockers
• Converting vague requirements into technical details
• Preparing issues for sprint planning

Core Capabilities

Decision Tree

Analyze Issue
├─ Epic? → Epic Decomposition
├─ Missing Criteria? → Acceptance Criteria Extraction
├─ Unclear Complexity? → Complexity Analysis
├─ Unknown Dependencies? → Dependency Mapping
├─ Unclear Technical Approach? → Technical Design
├─ Missing Tests? → Test Case Generation
└─ Sprint Context Needed? → Sprint Context Analysis

Epic Decomposition

When: Epic type, vague requirements, multi-sprint work

Process:

1. Extract business objectives and success metrics
2. Identify user journeys and personas
3. Break into stories with acceptance criteria
4. Map story dependencies
5. Create phased roadmap

Example Output:

# Epic Decomposition: EPIC-001

## Business Objectives
- Primary goal: [What problem solved?]
- Success metrics: [ROI/impact?]

## Stories by Phase
### Phase 1: Foundation (Sprint 1)
- STORY-001: [Title] (5 pts) → Depends on [deps]
- STORY-002: [Title] (3 pts)

### Phase 2: Core (Sprint 2-3)
- STORY-003: [Title] (8 pts) → Depends on STORY-001

## Dependency Graph
STORY-001 → STORY-003 → STORY-005

## Total Effort
- Points: 24 | Sprints: 4 | Team: 2-3 devs

Acceptance Criteria Extraction

When: Vague/missing criteria, implicit requirements

Extraction Patterns:

• Modal verbs: must, should, shall, will, requires, needs
• Behavior keywords: when, if, then, after, before, given
• Constraints: only, except, within X time, at least, max

Example Output:

# Acceptance Criteria: ISSUE-001

## Extracted Criteria

### Functional
- [ ] AC1: User can submit form when all required fields filled
- [ ] AC2: System sends confirmation email within 1 minute

### Non-Functional
- [ ] AC3: Form submission < 2 seconds
- [ ] AC4: Mobile support (iOS 14+, Android 10+)

### Security
- [ ] AC5: TLS 1.3 encryption in transit
- [ ] AC6: Rate limit: 5 submissions/hour per user

## Test Scenarios

Given user on contact form When user enters valid data Then form submits successfully And confirmation email sent

## Missing Info (Needs Clarification)

- Email service failure behavior?
- Max character limits?
- Duplicate submission handling?

Complexity Analysis

When: Estimate missing/unclear, assess technical risk

Complexity Scoring:

Story Point Mapping:

• 1-10 weighted score → 1 pt (trivial)
• 11-20 → 2 pts (simple)
• 21-30 → 3 pts (moderate)
• 31-40 → 5 pts (complex)
• 41-50 → 8 pts (very complex)
• 51+ → 13 pts or break down

Example Output:

# Complexity: ISSUE-001

## Summary
- **Points:** 5 | **Confidence:** Medium (70%)
- **Risk:** Medium | **Duration:** 2-3 days

## Code Impact
- Files: 8 | LOC: 300-400 | New Files: 2-3
- Integration Points: Auth0, DB, 3 services
- Risk Factors: Security, DB migration

## Historical Comparison
- Similar issue PROJ-234 (8 pts, 4 days)
- Similar issue PROJ-189 (3 pts, 1 day)
- Team velocity: 25 pts/sprint | 2.5 days per 5-pt story

## Scoring
- Code Changes: 3/5 × 0.25 = 0.75
- Integration: 3/5 × 0.20 = 0.60
- Risk: 4/5 × 0.20 = 0.80
- Testing: 3/5 × 0.15 = 0.45
- Dependencies: 2/5 × 0.10 = 0.20
- Uncertainty: 3/5 × 0.10 = 0.30
- **Total: 3.10 → 5 Points**

## Recommended Actions Before Starting
- [ ] Clarify unclear requirements
- [ ] Define performance SLA
- [ ] Create rollback plan
- [ ] Security review acceptance criteria

Dependency Mapping

When: Complex issues, cross-team work, risk assessment

Dependency Types:

1. Jira Links: Blocks/blocked by, parent/child, related
2. Code: Shared libs, API contracts, DB schemas, config
3. Team: Other team's work, shared resources, reviews
4. External: Third-party APIs, infrastructure, compliance

Example Output:

# Dependency Analysis: ISSUE-001

## Summary
- Total: 7 | Blocking: 2 (CRITICAL) | Code: 3 | Team: 2 | External: 1

## Critical Path (5 days)
START → PROJ-100 (Auth API) → PROJ-123 (Current) → PROJ-124 (Frontend) → END

## Blocking Issues
### PROJ-100: Auth0 API Configuration
- Status: In Progress | ETA: 2 days
- Impact: Cannot start until credentials ready
- Action: Daily follow-up

### PROJ-111: Database Migration Framework
- Status: In Review | ETA: 1 day
- Impact: Need migration CLI for schema changes
- Action: Review and test PR locally

## Code Dependencies
- auth.service.ts: Stable, low risk
- user.model.ts: Active dev (PROJ-98), medium risk
- Auth0 API: Well-documented, low risk

## Team Dependencies
- Security Review (2-3 days after PR)
- API Documentation (1 day after merge)

## Risk Matrix
| Dependency | Type | Status | Risk | Mitigation |
|-----------|------|--------|------|-----------|
| PROJ-100 | Blocking | In Progress | High | Daily follow-up |
| PROJ-111 | Blocking | In Review | Medium | Review PR |
| Security | Team | Pending | Medium | Schedule early |

## Execution Order
1. **Pre-work (Days 1-2):** Wait for blockers, test migration, review docs
2. **Implementation (Days 3-4):** Coordinate, implement, write tests
3. **Review (Day 5):** Security, PR, deploy to staging

## Parallel Opportunities
- Write comprehensive tests (no blockers)
- Draft documentation (no blockers)
- Design API spec (no blockers)

Technical Design Enhancement

When: Unclear approach, need architecture guidance, identify components

Key Areas:

• Architecture impact diagram (mermaid)
• Proposed code changes (1-2 examples)
• Database schema changes
• API endpoints
• Implementation patterns (token rotation, graceful degradation, security)
• Risk mitigation strategies
• Performance considerations and targets
• Testing strategy (unit, integration, security, performance)
• Rollout plan (dev → staging → prod with gradual rollout)
• Monitoring metrics and alerts

Integration with MCP Tools

const issue = await mcp.atlassian.getIssue(issueKey);
const criteria = extractAcceptanceCriteria(issue.fields.description);
const similarIssues = await mcp.atlassian.searchIssues(
  `text ~ "${issue.fields.summary}" AND status = Done`
);
const blockers = issue.fields.issuelinks.filter(l => l.type === 'Blocks');
await mcp.atlassian.updateIssue(issueKey, {
  fields: {
    customfield_10100: criteria,
    customfield_10101: estimatedPoints,
    description: enhancedDescription
  }
});

Enrichment Automation

Automatic Triggers:

• Issue Created → Extract criteria, find similar issues, suggest estimate
• Ready for Development → Full dependency analysis, design suggestions, tests
• Assigned to Sprint → Capacity check, complexity re-assessment, resources
• Description Updated → Re-extract criteria, update estimate, flag risks

Manual Commands:

/enrich-task PROJ-123                    # Full enrichment
/enrich-task PROJ-123 --criteria-only   # Criteria only
/enrich-task PROJ-123 --complexity-only # Complexity only
/enrich-sprint "Sprint 24"               # Batch enrich sprint

Best Practices

1. Progressive Enrichment: Start lightweight, add detail as issue progresses
2. Human-in-the-Loop: Flag uncertain analysis, provide confidence scores
3. Context Preservation: Link to sources, document reasoning
4. Team Adaptation: Learn from estimation accuracy, adapt to velocity
5. Continuous Learning: Track accuracy, refine models, update patterns

Example: Enriching Vague Bug

Before:

Title: Login not working
Description: Users can't log in sometimes

After:

## Missing Info (Critical)
- Reproduction steps?
- Affected users (all or subset)?
- Frequency (always, sometimes, rarely)?
- Browser/device specifics?
- Error messages?
- Correlation with deployments?

## Similar Issues
- PROJ-89: "Login timeout issues" (3 pts, 1 day)
  Root cause: Session expiry misconfiguration

## Investigation Steps
1. Check logs for auth errors
2. Review Auth0 dashboard
3. Test in different browsers
4. Check correlation with time periods
5. Review auth.service.ts changes

Cannot Estimate Without Reproduction → Move to "Needs More Info"

Summary

This skill enables comprehensive task enrichment through:

• Intelligent Extraction: Auto-parse requirements from descriptions
• Complexity Assessment: Technical analysis with accurate estimates
• Dependency Mapping: Identify all blockers before work starts
• Design Guidance: Suggest architecture and patterns
• Historical Learning: Leverage past issues
• MCP Integration: Seamless Jira/Confluence automation

Use in EXPLORE phase to ensure comprehensive context before development.