name: orchestrate description: Wire Commands, Agents, and Skills together for complex features. Use when building features that need research, planning, and implementation phases. user-invocable: true

Orchestrate - Multi-Phase Feature Development

Build features through structured phases with validation gates.

The Pattern

/develop <feature>
  │
  ├── Phase 1: Research (orchestrator agent)
  │   └── Score confidence → GO/HOLD
  │
  ├── Phase 2: Plan (orchestrator agent)
  │   └── Present plan → wait for approval
  │
  ├── Phase 3: Implement (orchestrator agent)
  │   └── Execute plan → quality gates
  │
  └── Phase 4: Review (reviewer agent)
      └── Code review → commit

Usage

When asked to build a feature:

1. Start with research: Delegate to the orchestrator agent or scout agent to explore the codebase
2. Wait for GO/HOLD: Don't proceed if confidence is below 70
3. Present a plan: List all files to change, the approach, and risks
4. Get approval: Never implement without explicit "proceed"
5. Implement step by step: Quality gates every 5 edits
6. Review before commit: Run the reviewer agent on changes

When to Use This

• Feature touches >5 files
• Architecture decisions needed
• Requirements are unclear or complex
• Cross-cutting concerns (auth, logging, error handling)
• New patterns not yet established in the codebase

When NOT to Use This

• Quick bug fixes (just fix it)
• Single-file changes
• Well-understood patterns (follow existing code)
• Documentation-only changes

Agent Selection

| Phase | Agent | Why | |-------|-------|-----| | Research | scout (background, worktree) | Non-blocking exploration | | Plan | orchestrator (opus, memory) | Deep reasoning, pattern recall | | Implement | orchestrator (opus, memory) | Full tool access | | Review | reviewer (read + bash) | Security and quality focus | | Debug | debugger (opus, memory) | Systematic investigation |

Integration with Pro-Workflow

• Corrections during implementation trigger self-correction loop
• Quality gates fire at checkpoints via hooks
• Learnings are captured at the end of each phase
• Session handoff works across phases