feat: Duck Ecosystem - Multi-discipline review with specialized duck personas

[nesquikm]

🦆 Duck Enhancement Proposal

💡 The Problem

Getting a comprehensive review requires expertise across multiple disciplines. Currently you'd need to manually prompt for security, performance, UX, ops perspectives separately.

🚀 Proposed Solution

duck_ecosystem({
  prompt: "Review this API design: POST /users { name, email, password }",
  species: ["security", "performance", "ux", "ops", "architecture"],  // or "all"
  synthesizer: "openai"  // provider for final synthesis
})

// Returns
{
  reviews: [
    {
      species: "security",
      persona: "SecDuck 🔒",
      findings: [
        { severity: "high", issue: "Password in plain request body", recommendation: "Use HTTPS, hash immediately" },
        { severity: "medium", issue: "No rate limiting mentioned", recommendation: "Add rate limits" }
      ]
    },
    {
      species: "performance",
      persona: "PerfDuck ⚡",
      findings: [
        { severity: "low", issue: "Email validation on every request", recommendation: "Cache validation results" }
      ]
    },
    // ... more species
  ],
  synthesis: "Overall the API design is functional but has critical security gaps...",
  priority_actions: ["1. Add HTTPS requirement", "2. Hash passwords", "3. Add rate limiting"]
}

🦆 Species Definitions

• SecDuck 🔒 - Security vulnerabilities, auth, injection, OWASP
• PerfDuck ⚡ - Latency, throughput, caching, N+1 queries
• UXDuck 🎨 - API ergonomics, error messages, discoverability
• OpsDuck 🔧 - Monitoring, deployment, failure modes, SLAs
• ArchDuck 🏗️ - Scalability, maintainability, coupling, patterns

📋 Implementation

1. src/tools/duck-ecosystem.ts - Main tool
2. src/ecosystem/species.ts - Persona definitions with system prompts
3. Parallel execution across all species
4. Synthesis duck combines findings with priority ranking

🌟 Research Backing

• Specialized AI agents
• Code review subagents