Qt Code Review Skills are AI-powered agent tools for automated code quality analysis of Qt6 C++ and QML codebases. The workflow runs a deterministic Python linter (60+ C++ rules, 47+ QML checks) followed by six parallel deep-analysis agents covering areas like memory ownership, thread safety, binding loops, and performance anti-patterns. Only findings above an 80/100 confidence threshold are reported as confirmed issues. The skills are read-only, require Python 3 and optionally Git or qmllint, and have been tested with Claude Code CLI, GitHub Copilot, and frontier models like GPT and Gemini. Available via a GitHub repository and soon as a Claude marketplace plugin.
Sort: