/coordinate — Advanced Multi-Agent Coordination

$ARGUMENTS

---

🔴 CRITICAL RULES

1. Load coordinator-mode skill — Read .agent/skills/coordinator-mode/SKILL.md first
2. Phase-based execution — Research → Synthesis → Implementation → Verification
3. Never delegate understanding — Write specific prompts, not vague instructions
4. Parallel reads, sequential writes — Read-only tasks can run in parallel

---

Task

Use the orchestrator agent with this context:

CONTEXT:
- User Request: $ARGUMENTS
- Mode: COORDINATOR (advanced orchestration)
- Skill: Load coordinator-mode skill for patterns

WORKFLOW:
1. DECOMPOSE the request into worker subtasks
2. CLASSIFY each subtask: Research | Implementation | Verification
3. DISPATCH workers (parallel for reads, sequential for writes)
4. SYNTHESIZE results — don't copy-paste, add insight
5. VERIFY completeness before reporting to user

RULES:
1. Follow coordinator-mode/SKILL.md protocol
2. Brief workers like smart colleagues — full context, specific scope
3. Never write "based on your findings, fix it" — prove YOU understood
4. Start with 2-3 workers, add more after synthesis if needed
5. Research before implementation — ALWAYS

---

Expected Output

Deliverable	Description
Worker Dispatch	List of agents invoked with task descriptions
Synthesis Report	Consolidated findings with YOUR analysis
Action Items	Specific next steps with file paths

---

After Coordination

[OK] Coordination complete

Agents dispatched: [count]
Phases completed: Research → Synthesis → [Implementation] → [Verification]

Key findings:
- [Finding 1]
- [Finding 2]

Next steps:
- [ ] [Action item 1]
- [ ] [Action item 2]

---

Generated by Agent Bridge