What problem does it solve? Reviewing NetEase Mod code involves many dimensions—Python 2.7 compatibility, encoding safety, API correctness, UI bindings, RPC communication, configuration management, and performance—and checking them all manually is error-prone and inconsistent. This Skill acts as a central review orchestrator that classifies each change, dispatches the right specialized review sub-skills in the correct order, and merges their findings into one unified report. ## Core Features & Use Cases - Layered Review Pipeline: Runs blocking checks first (Py2.7 syntax, encoding, API verification), then design and convention review, then domain-specific checks (UI, RPC, config), then simplification and performance, and finally runtime instrumentation when static analysis is inconclusive. - Adaptive Review Levels: Automatically selects quick, standard, or deep review based on change size, or honors an explicitly requested level. - Unified Severity Report: Produces a consolidated report with 🔴/🟠/🟡/🟢 severity tiers, per-skill issue counts, and a pass/fail gate (🔴 = 0 and 🟠 ≤ 2). - Use Case: After finishing a feature that touches JSON UI and RPC endpoints, ask for a deep review; the orchestrator will invoke the syntax, clean-code, UI-binding, and RPC sub-skills in parallel, deduplicate findings, and return a single actionable report. ## Quick Start Review my current changes with a deep code review and give me a consolidated severity-ranked report.
