skills/crew/workflow/SKILL.md
Core wicked-crew v3 workflow engine with capability-based orchestration and phase progression. Drives projects through crew phases: clarify → design → test-strategy → build → test → review. Smart decisioning analyzes signals to determine which specialists to engage per phase. Use when: "crew phases", "phase plan", "workflow execution", "start a project", "clarify outcome", "design phase", "build phase", "approve phase", "crew workflow", "phase progression", "QE gate", "shift-left testing", or structured delivery guidance.
npx skillsauth add mikeparcewski/wicked-garden workflowInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Capability-based orchestration with smart specialist engagement.
User Input → Smart Decisioning → Specialist Discovery → Engagement
│ │
├── Signal Detection ├── Available specialists
├── Complexity Score ├── Personas
└── Ambiguity Check └── Fallback agents
clarify → design → test-strategy → build → test → review
Each phase has: Clear deliverables, specialist engagement based on signals, mandatory quality gate.
Gates are hard enforcement — not advisory. phases.json defines per-phase:
min_gate_score (0.6-0.8): gates below threshold block advancementvalid_skip_reasons: enumerated skip reasons; free-text rejectedskip_complexity_threshold: prevents skipping at high complexity (e.g., test-strategy at >= 3)Gate verdicts: APPROVE → proceed. CONDITIONAL → conditions-manifest.json written, must verify before next phase. REJECT → blocks advancement, mandatory rework.
CONDITIONAL auto-resolution (AC-4.4): spec gaps fixed inline; intent changes escalate to user/council.
Build depends on design (depends_on: ["clarify", "design"]). Rollback: CREW_GATE_ENFORCEMENT=legacy.
| Signal | Keywords | Specialists | |--------|----------|-------------| | Security | auth, encrypt, token, jwt | platform, compliance | | Performance | scale, optimize, cache | engineering, qe | | Product | user, feature, story | product | | Compliance | SOC2, HIPAA, audit | platform | | Ambiguity | maybe, should we, options | jam | | Complexity | integration, migrate, refactor | delivery, engineering |
Crew discovers specialists via specialist.json in each plugin.
| Specialist | Fallback | |------------|----------| | jam | facilitator | | qe, product | reviewer | | engineering (arch) | researcher | | engineering, platform | implementer |
Goal: Define success criteria Deliverables: Outcome statement, success criteria, scope boundaries Specialists: jam (if ambiguous), product
Goal: Architect the solution Deliverables: Architecture docs, pattern identification, technical approach Specialists: product, engineering
Goal: Define test strategy before building Deliverables: Test scenarios, risk assessment, edge cases Specialists: qe, platform (if security signals)
Goal: Implement the solution Deliverables: Working implementation, progress tracking, tests passing Specialists: engineering, platform (if infra)
Goal: Multi-perspective validation Deliverables: Review findings, recommendations, sign-off Specialists: product, qe, platform (if regulated)
Projects can be archived when complete or paused:
sh "${CLAUDE_PLUGIN_ROOT}/scripts/_python.sh" "${CLAUDE_PLUGIN_ROOT}/scripts/crew/crew.py" archive <name> # Sets archived=true
sh "${CLAUDE_PLUGIN_ROOT}/scripts/_python.sh" "${CLAUDE_PLUGIN_ROOT}/scripts/crew/crew.py" unarchive <name> # Reactivates
sh "${CLAUDE_PLUGIN_ROOT}/scripts/_python.sh" "${CLAUDE_PLUGIN_ROOT}/scripts/crew/crew.py" list # Excludes archived by default
sh "${CLAUDE_PLUGIN_ROOT}/scripts/_python.sh" "${CLAUDE_PLUGIN_ROOT}/scripts/crew/crew.py" list --include-archived
Phase operations are blocked on archived projects.
| Command | Purpose |
|---------|---------|
| /wicked-garden:crew:start | Begin project with smart decisioning |
| /wicked-garden:crew:status | View current state and engaged specialists |
| /wicked-garden:crew:execute | Run current phase with specialists |
| /wicked-garden:crew:approve | Approve and advance phase |
| /wicked-garden:crew:just-finish | Autonomous completion |
| /wicked-garden:crew:gate | Run QE gate (value/strategy/execution) |
| /wicked-garden:crew:evidence | Query evidence for a task |
crew:project:started:success
crew:specialist:engaged:success
crew:specialist:unavailable:warning (with fallback)
crew:phase:started:success
crew:phase:completed:success
crew:phase:approved:success
| Plugin | Enhancement | Fallback | |--------|-------------|----------| | wicked-garden:kanban | Persistent task board | TodoWrite | | wicked-garden:mem | Cross-session learning | Project files |
See Integration Details for usage patterns.
Gate decisions and artifacts are tracked as evidence. See Evidence Tracking for:
Customize in the wicked-crew local storage domain as config.yaml:
defaults:
always_engage: [qe] # Always use QE specialist
complexity_threshold: 4 # Lower = more specialists
development
--- name: large-scale-migration description: How to execute a LARGE MECHANICAL change across any codebase with LEVERAGE instead of an agent-grind or hand-edits — a cross-cutting migration, refactor, rename, dialect/framework/DB port, library adoption, or bulk transform. The map→transform→gate pattern: a deterministic transform driven by a source-of-truth map, proven by a differential-equivalence gate. Use when the work is "migrate all X to Y", "rename Z everywhere", "port to a new DB/dialect/fra
testing
v11 LLM-based work-shape classifier. Replaces the regex archetype detector with the model's own reasoning. Reads the user's prompt, picks the right archetype(s) from the catalog, identifies signals (blast_radius, novelty, reversibility, etc.), and persists to SessionState so subsequent turns steer correctly. Use when: the prompt_submit hook emitted a `<wg classify-due />` directive, OR explicitly invoked at session start, OR when re-classifying after the user changes scope mid-session.
tools
v11 work-shape archetype runner. When a prompt has been routed to one of the 9 archetypes (triage, explore, specify, decide, ship, review, incident, build, migrate), this skill is the entry point. It picks the right per-archetype playbook from refs/ and executes the phase shape declared in `.claude-plugin/archetypes.json`. Use when: a `<wg archetype="X">` or `<wg archetypes>` system-reminder tag appears, an explicit "let's run the X archetype" request, or when one of the per-archetype slash commands resolves to this skill.
development
Show or set the session intent variable. Intent gates how loud the framework is — simple-edit (silent), feature/research (synthesis directive), rigor (full crew context). Auto-detected on turn 1; this skill overrides explicitly. Sticky for the session. Use when: "set intent", "intent override", "/wicked-garden:intent", "make the framework quiet", "force rigor", "what's my intent".