Workflow Skill (v3)
Capability-based orchestration with smart specialist engagement.
v3 Architecture
User Input → Smart Decisioning → Specialist Discovery → Engagement
│ │
├── Signal Detection ├── Available specialists
├── Complexity Score ├── Personas
└── Ambiguity Check └── Fallback agents
Phase Progression
clarify → design → qe → build → review
Each phase has: Clear deliverables, specialist engagement based on signals, approval gate.
Smart Decisioning
Signal Categories
| Signal | Keywords | Specialists |
|---|---|---|
| Security | auth, encrypt, token, jwt | devsecops, compliance |
| Performance | scale, optimize, cache | appeng, qe |
| Product | user, feature, story | product, strategy |
| Compliance | SOC2, HIPAA, audit | compliance |
| Ambiguity | maybe, should we, options | jam |
| Complexity | integration, migrate, refactor | pmo, arch |
Complexity Scoring (0-7)
- •0-2: Simple → Built-in agents only
- •3-4: Moderate → Core specialists
- •5-7: Complex → All relevant + PMO
Specialist Engagement
Discovery
Crew discovers specialists via specialist.json in each plugin.
Fallback Agents
| Specialist | Fallback |
|---|---|
| jam | facilitator |
| qe, strategy | reviewer |
| arch | researcher |
| appeng, devsecops | implementer |
Phase Details
Clarify
Goal: Define success criteria Deliverables: Outcome statement, success criteria, scope boundaries Specialists: jam (if ambiguous), product
Design
Goal: Architect the solution Deliverables: Architecture docs, pattern identification, technical approach Specialists: strategy, arch, appeng
QE (Quality Engineering)
Goal: Define test strategy before building Deliverables: Test scenarios, risk assessment, edge cases Specialists: qe, devsecops (if security signals)
Build
Goal: Implement the solution Deliverables: Working implementation, progress tracking, tests passing Specialists: appeng, devsecops (if infra)
Review
Goal: Multi-perspective validation Deliverables: Review findings, recommendations, sign-off Specialists: strategy, qe, compliance (if regulated)
Commands Reference
| Command | Purpose |
|---|---|
/wicked-crew:start | Begin project with smart decisioning |
/wicked-crew:status | View current state and engaged specialists |
/wicked-crew:execute | Run current phase with specialists |
/wicked-crew:approve | Approve and advance phase |
/wicked-crew:just-finish | Autonomous completion |
/wicked-crew:gate | Run QE gate (value/strategy/execution) |
/wicked-crew:evidence | Query evidence for a task |
Event Flow
Project Start
crew:project:started:success crew:specialist:engaged:success crew:specialist:unavailable:warning (with fallback)
Phase Transitions
crew:phase:started:success crew:phase:completed:success crew:phase:approved:success
Utility Integration
| Plugin | Enhancement | Fallback |
|---|---|---|
| wicked-kanban | Persistent task board | TodoWrite |
| wicked-mem | Cross-session learning | Project files |
| wicked-cache | Cached analysis | Re-run each time |
See Integration Details for usage patterns.
Evidence Tracking
Gate decisions and artifacts are tracked as evidence. See EVIDENCE.md for:
- •Evidence tiers (L1-L4)
- •Artifact naming conventions
- •Automatic vs manual evidence collection
Configuration
Customize in ~/.something-wicked/wicked-crew/config.yaml:
defaults: always_engage: [qe] # Always use QE specialist complexity_threshold: 4 # Lower = more specialists