Qt Qml Review

This is a comprehensive QML code review system that runs 47+ deterministic lint rules in Python, then spins up six parallel analysis agents to catch the subtle stuff: binding loops, layout feedback cycles, memory leaks from dynamic object creation, and delegate reuse bugs. It's read-only and reports only high-confidence issues above 80/100. You'd run this before committing QML or when auditing a codebase. The interesting part is the two-phase design: fast mechanical checks first, then deep semantic analysis that actually traces symbol references and checks visual tree relationships. It also wraps system qmllint if available. Honestly a serious piece of work for teams shipping non-trivial Qt6 UIs who are tired of runtime surprises.

Install to Claude Code

npx -y skills add theqtcompanyrnd/agent-skills --skill qt-qml-review --agent claude-code