Use Case Guide
Find the right SpecWeave skill or command for what you're trying to do. Every skill supports three invocation methods:
"Let's build a checkout flow"
"I want to..." Quick Lookup
Planning & Starting Work
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Start a new feature | "Let's build X" | sw:increment "X" | increment "X" |
| Get help writing specs | "Write specs" | sw:pm | pm |
| Design system architecture | "Design the system" | sw:architect | architect |
| Plan a product roadmap | "Plan the roadmap" | sw:roadmap-planner | roadmap-planner |
| Break down a complex feature | "Coordinate all agents" | sw:role-orchestrator | role-orchestrator |
| Create tasks with test plans | "Generate tasks" | sw:plan | plan |
Implementing Features
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Work autonomously (hours) | "Ship while I sleep" | sw:auto | auto |
| Work task-by-task manually | "Start implementing" | sw:do | do |
| Work with multiple agents | "Parallel agents" | sw:auto-parallel | auto-parallel |
| Orchestrate parallel teams | "Team work on X" | sw:team-lead "X" | team-lead "X" |
| Brainstorm with agents | "Brainstorm X" | sw:team-lead "brainstorm X" | team-lead "brainstorm X" |
| Plan with PM + Architect | "Plan X with a team" | sw:team-lead "plan X" | team-lead "plan X" |
| Research a topic with agents | "Research X" | sw:team-lead "research X" | team-lead "research X" |
| Check auto mode progress | "Check auto progress" | sw:auto-status | auto-status |
| Get frontend help | "Help with React" | sw:architect | architect |
| Get Node.js API help | "Node.js API" | /backend:nodejs | nodejs |
| Optimize database queries | "Optimize queries" | /backend:database-optimizer | database-optimizer |
| Set up Kubernetes | "K8s deployment" | /k8s:deployment-generate | deployment-generate |
| Design Kafka architecture | "Kafka event design" | /kafka:architect | kafka-architect |
Testing & Quality
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Validate quickly (rules) | "Check quality" | sw:validate | validate |
| AI quality assessment | "Assess quality" | sw:qa --gate | qa --gate |
| Deep implementation audit | "Review my work" | sw:grill | grill |
| Follow TDD strictly | "Test-driven development" | sw:tdd-cycle | tdd-cycle |
| Write failing tests first | "Write failing tests" | sw:tdd-red | tdd-red |
| Write E2E tests | "E2E tests" | sw:e2e | e2e |
| Code review my changes | "Review code" | sw:code-reviewer | code-reviewer |
| Review a specific PR | "Review PR 42" | sw:code-reviewer --pr 42 | code-reviewer --pr 42 |
Completing & Managing Work
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Finish and get next task | "What's next?" | sw:next | next |
| Close specific increment | "We're done" | sw:done 0007 | done 0007 |
| Check progress | "How far along?" | sw:progress | progress |
| Pause for later | "Put this on hold" | sw:pause | pause |
| Resume paused work | "Continue working" | sw:resume | resume |
| Abandon increment | "Cancel this" | sw:abandon | abandon |
Syncing with External Tools
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Sync to GitHub Issues | "Sync to GitHub" | sw-github:sync | github-sync |
| Sync to JIRA | "Sync to JIRA" | sw-jira:sync | jira-sync |
| Sync to Azure DevOps | "Sync to ADO" | sw-ado:sync | ado-sync |
| Configure external sync | "Set up sync" | sw:external-sync-wizard | external-sync-wizard |
Documentation
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Write technical docs | "Write documentation" | sw:docs-writer | docs-writer |
| Update living docs | "Update the docs" | sw:sync-docs | sync-docs |
| Navigate project docs | "Show me the docs" | sw:living-docs-navigator | living-docs-navigator |
| Build Docusaurus site | "Docusaurus setup" | /docs:docusaurus | docusaurus |
Security & Compliance
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Security assessment | "Security review" | sw:security | security |
| Detect security patterns | "Check for vulnerabilities" | sw:security-patterns | security-patterns |
| SOC 2/HIPAA compliance | "Compliance check" | sw:compliance-architecture | compliance-architecture |
| PCI-DSS for payments | "PCI compliance" | /payments:pci-compliance | pci-compliance |
Cost & Performance
| I want to... | Natural Language | Claude Code | Other AI |
|---|---|---|---|
| Optimize cloud costs | "Reduce cloud costs" | /cost:cost-optimization | cost-optimization |
| Analyze AWS costs | "AWS cost analysis" | /cost:aws-cost-expert | aws-cost-expert |
| Improve performance | "Optimize performance" | sw:performance | performance |
By Role
I'm a Product Manager
sw:pm # Requirements, user stories, ACs
sw:roadmap-planner # Quarterly planning, prioritization
sw:increment "feature" # Start new work
sw:progress # Track team progress
I'm an Architect
sw:architect # System design, ADRs
sw:role-orchestrator # Coordinate multiple agents
/infra:devops # CI/CD, deployment
sw:security # Security review
I'm a Frontend Developer
sw:architect # React/Vue patterns
sw:architect # Next.js specifics
sw:architect # Component library
sw:auto # Autonomous implementation
I'm a Backend Developer
/backend:nodejs # Node.js APIs
/backend:database-optimizer # SQL optimization
/kafka:architect # Event-driven design
sw:auto # Autonomous implementation
I'm a DevOps Engineer
/infra:devops # CI/CD pipelines
/k8s:deployment-generate # K8s manifests
/k8s:gitops-workflow # ArgoCD/Flux
/infra:terraform # Terraform IaC
I'm a QA Engineer
/testing:qa # Test strategy
/testing:e2e # Playwright E2E
/testing:unit # Unit test patterns
sw:tdd-cycle # TDD workflow
sw:grill # Implementation audit
I'm a Security Engineer
sw:security # Vulnerability assessment
sw:security-patterns # Real-time detection
sw:compliance-architecture # Compliance frameworks
/payments:pci-compliance # PCI-DSS
By Phase
Phase 1: Planning
Commands/Skills:
sw:increment "feature"- Create incrementsw:pm- Product managementsw:architect- Architecture designsw:role-orchestrator- Multi-agent coordination
Phase 2: Implementation
Commands/Skills:
sw:auto- Autonomous executionsw:do- Manual executionsw:*- Frontend skills/backend:*- Backend skills
Phase 3: Quality
Commands/Skills:
sw:validate- Rule-based checkssw:qa --gate- AI quality gatesw:grill- Deep auditsw:code-reviewer- Code review
Phase 4: Completion
Commands/Skills:
sw:next- Complete and suggest nextsw:done- Close incrementsw-github:sync- Sync to GitHub
Common Workflows
Start to Finish (Autonomous)
# 1. Create increment
sw:increment "User authentication with JWT"
# 2. Let it run (go grab coffee, lunch, or sleep)
sw:auto
# 3. Check from another terminal (optional)
sw:auto-status
# 4. Complete when done
sw:next
# 5. Sync to GitHub (optional)
sw-github:sync 0007
TDD Workflow
# 1. Create increment
sw:increment "Payment processing"
# 2. TDD cycle
sw:tdd-cycle
# OR step by step:
sw:tdd-red # Write failing tests
sw:tdd-green # Minimal implementation
sw:tdd-refactor # Clean up
# 3. Quality gate
sw:qa --gate
# 4. Complete
sw:next
Quality-First Release
# 1. Quick validation
sw:validate 0007
# 2. Deep audit
sw:grill 0007 --full
# 3. AI quality gate
sw:qa 0007 --gate
# 4. Code review
sw:code-reviewer
# 5. Security check
sw:security
# 6. Complete
sw:done 0007
Multi-Repo Project
# 1. Detect repo structure
sw:umbrella-repo-detector
# 2. Create coordinated increment
sw:increment "Feature spanning FE + BE"
# 3. Parallel execution
sw:auto-parallel
# 4. Sync all to GitHub
sw-github:github-multi-project
Decision Trees
Which Execution Mode?
Start Implementation
│
▼
┌───────────────────┐
│ Do I need to make │
│ decisions during │──Yes──▶ sw:do (manual)
│ implementation? │
└───────────────────┘
│No
▼
┌───────────────────┐
│ Are there isolated│
│ parallel work │──Yes──▶ sw:auto-parallel
│ streams (FE/BE)? │
└───────────────────┘
│No
▼
sw:auto (autonomous)
Which Quality Check?
Want Quality Check
│
▼
┌───────────────────┐
│ Quick validation │
│ (rules only)? │──Yes──▶ sw:validate
└───────────────────┘
│No
▼
┌───────────────────┐
│ Deep code audit │
│ with parallel │──Yes──▶ sw:grill
│ analysis? │
└───────────────────┘
│No
▼
sw:qa --gate (AI quality)
Which Sync Command?
Sync to External Tool
│
▼
┌───────────────────┐
│ Which tool? │
└───────────────────┘
│
┌───┴───┬───────┐
▼ ▼ ▼
GitHub JIRA ADO
│ │ │
▼ ▼ ▼
sw-github sw-jira sw-ado
:sync :sync :sync
Skill Auto-Activation Keywords
Skills activate automatically when you use these keywords in natural language. You can also use slash commands or CLI keywords directly.
| Natural Language Keywords | Claude Code | Other AI |
|---|---|---|
| "user story", "acceptance criteria", "requirements" | sw:pm | pm |
| "architecture", "ADR", "design decision" | sw:architect | architect |
| "React", "Vue", "Angular", "frontend" | sw:architect | architect |
| "Node.js", "Express", "API endpoint" | /backend:nodejs | nodejs |
| "database", "SQL", "query optimization" | /backend:database-optimizer | database-optimizer |
| "Kubernetes", "K8s", "pods", "deployment" | /k8s:deployment-generate | deployment-generate |
| "Kafka", "events", "streaming" | /kafka:architect | kafka-architect |
| "test", "TDD", "unit test", "E2E" | sw:tdd-cycle | tdd-cycle |
| "security", "OWASP", "vulnerability" | sw:security | security |
| "compliance", "SOC 2", "HIPAA", "GDPR" | sw:compliance-architecture | compliance-architecture |
Next Steps
- Skills Reference - Complete skills list
- Commands Reference - Complete commands list
- Quick Start - Get started in 5 minutes