**기능 준비 상태 감사 프롬프트**

# Feature Readiness Audit **Feature:** ${featureName} **Auditor:** Senior Principal Engineer **Date:** ${currentDate} ## 1. Implementation Quality - **Strengths:** Clear separation of concerns; modular functions; consistent error handling patterns. - **Weaknesses:** Missing input validation on public API; some logic duplicated across modules; insufficient comments on complex state transitions. ## 2. Role & Dependencies - **Dependencies:** Relies on `authService`, `cacheLayer`, and `configStore`. No circular imports detected. - **Integration:** Correctly wired into main event bus; however, feature is tightly coupled to legacy `dataFetcher`, limiting testability. ## 3. Expected vs. Actual Behavior - **Expected:** Should return cached results within 50ms; fallback to DB on cache miss. - **Actual:** Cache hit rate is 72% (target ≥90%); p95 latency is 210ms due to serial DB calls. Behavior is functionally correct but performance misses spec. ## 4. Edge Cases & Risks - **Edge Cases:** Empty payloads handled; concurrent duplicate requests cause race condition (no idempotency key). - **Risks:** High memory usage under burst load (unbounded cache); silent failure on third-party timeout (no retry/backoff). ## 5. Cross-Cutting Concerns - **Performance:** Suboptimal DB query (N+1) in `getUserData`; cache invalidation is time-based only. - **Security:** No rate limiting on public endpoint; sensitive data logged in debug mode. - **Scalability:** Stateless service, but cache is in-memory - will not scale horizontally. - **Maintainability:** Test coverage at 58%; missing integration tests for failure paths. ## 6. Readiness Score: **6/10** Justification: Functionally complete and integrated, but performance, concurrency safety, and scalability gaps prevent production readiness. Security hardening is also required. --- ## Next-Step Recommendations 1. **Immediate:** Add idempotency keys and request-level rate limiting. 2. **Short-term:** Optimize DB queries (eager loading), implement cache eviction policy (LRU), and add retry with exponential backoff. 3. **Medium-term:** Replace in-memory cache with Redis; add distributed tracing. 4. **Before release:** Increase test coverage to ≥80%, remove sensitive logging, and run load test at 2x expected peak. **Verdict:** Not ready for production. Re-audit after addressing items 1–3.