skills.rest
TypeThe0ry
TypeThe0ryCOMMUNITY
Explore all skills by this author →
GitHub

qt-cpp-review

Reviews Qt6 C++ code using deterministic linting and six parallel deep-analysis agents.

1|Updated Jul 12, 2026
Search Tags:#code quality#static analysis#code review#linting#c++#qt6#qabstractitemmodel
One-click install
npx skills add https://github.com/TypeThe0ry/Multi-Detect --skill qt-cpp-review-typethe0ry
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: qt-cpp-review
Source: https://github.com/TypeThe0ry/Multi-Detect/tree/main/ground-control/MultiDetectGCS/tools/skills/qt-cpp-review
Command: npx skills add https://github.com/TypeThe0ry/Multi-Detect --skill qt-cpp-review-typethe0ry

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Reviewing Qt6 C++ code by hand is slow and error-prone: model contract violations, ownership leaks, thread-safety bugs, and deprecated API usage are easy to miss. This Skill automates the review with a deterministic 60+ rule linter plus six focused analysis agents, reporting only high-confidence findings with mitigations. ## Core Features & Use Cases - Deterministic Linting: A single-pass Python linter checks 60+ rules covering includes, deprecated classes, model contracts, error handling, lifecycle, naming, and performance in under a second. - Six Parallel Deep-Analysis Agents: Focused agents cover model contracts, ownership and lifecycle, thread safety, API correctness, error handling, and performance, reporting only findings above 80/100 confidence. - Framework Mode: Optional Qt module-development rules (binary compatibility, export macros, d-pointers, qdoc, QML versioning) activated via argument or auto-detection. - Use Case: Before committing changes to a Qt6 ground-station application, run the review on the staged diff to catch a missing deleteLater() on a QNetworkReply and an unbalanced beginInsertRows/endInsertRows pair, then receive a consolidated report with file paths, line numbers, traces, and mitigations. ## Quick Start Ask the assistant to review your Qt6 C++ changes before committing, for example by saying "review my staged Qt C++ changes".

Frequently Asked Questions about qt-cpp-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review Qt6 C++ code before committing?▼

Invoke the review on your staged or unstaged changes and it runs a 60+ rule Python linter followed by six parallel analysis agents. The output is a consolidated report with file paths, line numbers, confidence scores, and prose mitigations, and it never modifies your code.

What does the Qt C++ code review check for?▼

It checks model contract violations (begin/end signal balance, dataChanged roles), ownership and lifecycle issues (missing deleteLater, Q_ASSERT side effects), thread safety, API naming, error handling (unchecked QFile::open, missing SSL handling), and performance anti-patterns like QRegularExpression in loops.

Does the Qt review skill support Qt framework module development?▼

Yes, passing the framework argument enables additional rules for binary compatibility, export macros, d-pointers, qdoc, and QML versioning. It can also auto-detect framework signals like Q_*_EXPORT macros or QT_BEGIN_NAMESPACE and suggest enabling framework mode.

What are the requirements to run the Qt review linter?▼

The linter requires Python 3.6 or later with no external dependencies. If Python is unavailable, the skill warns you and skips to the agent-driven deep analysis phase instead of failing.

Can the Qt code review modify my source files?▼

No, the review is strictly read-only. Both the linter and all six analysis agents only read files and report findings; mitigations are given in prose with no code patches applied.

Why are some review findings listed as investigation targets?▼

Findings with 60-79 confidence that cannot be fully verified, such as noexcept correctness requiring whole-program analysis, are listed separately as investigation targets. A maximum of 10 are reported, each with a specific verification action for a human reviewer.

Related Skills

Explore More Software Engineering Skills→
t-miura-024t-miura-024

mt-check-git-diff

Analyze git repository changes and retrieve detailed diffs for modified files.

Community
Intermediate
PioneerAIAcademyPioneerAIAcademy
3

review

Reviews pull requests by diffing against main and verifying findings with real test suites.

Official
Advanced
sek788432sek788432
6

cpp-static-analysis

Runs and interprets clang-tidy, scan-build, cppcheck, IWYU, and sanitizer checks for C++ code.

Community
Intermediate
RektooooooRektoooooo
2

swift-architecture

Automate Swift code reviews and architecture analysis with static analysis tools.

Community
Intermediate