/review

Use the pr-review-coordinator agent to conduct a comprehensive parallel review of the current PR.

The agent will:

1. Launch Jordan agent for architecture and production readiness assessment
2. Launch CodeRabbit for implementation details and security scanning
3. Synthesize results into actionable brief for human reviewer
4. Generate trust/verify/fix buckets with readiness score
5. Calculate confidence level for merge readiness
6. Track review metadata for learning patterns

After the review completes, you'll get a structured brief with categorized feedback:

• ✅ GREEN LIGHTS: Architecture validated, no security issues
• ⚠️ YELLOW LIGHTS: Business logic verification needed
• 🔴 RED LIGHTS: Issues that must be fixed

The agent will then present categorized feedback by topic and ask: "I have feedback in these categories: [Architecture, Security, Implementation, Testing]. Would you like to work through anything specific first?"

Options to follow up:

• Review a specific category in detail
• Show RED LIGHT (must-fix) issues
• Walk through YELLOW LIGHT (verification) topics
• Request human review with summary
• Handle manually with full findings

Context: $ARGUMENTS