accessibility
FreeEnsure accessibility in UI components including semantic HTML, ARIA attributes, keyboard navigation, and WCAG 2.2 AA compliance.
The English summary is being prepared. Showing the original description.
PyQt6 desktop GUI development rules -- signal/slot architecture, QSS theming, QThread concurrency, layout management, and cross-platform rendering. Enforces MVC separation and responsive UI patterns.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
This skill enforces rules for building production-quality PyQt6 desktop applications. The core principles are: strict MVC separation via signals/slots, never blocking the UI thread, centralized theming via QSS, and layout-manager-driven responsive design. These rules prevent the most common PyQt6 failures: frozen UIs, untestable coupling, and platform-specific rendering bugs.
| Anti-Pattern | Why It Fails | Correct Approach |
|---|---|---|
| Calling business logic directly from UI slots | Couples UI to logic; makes testing impossible and breaks MVC architecture | Emit signals from UI; connect to controller/service methods via slot |
| Running network or file I/O on the main thread | Blocks the Qt event loop; UI freezes until operation completes | Use QThread, QRunnable, or asyncio with qasync for background operations |
| Hardcoding pixel sizes and positions | Breaks on high-DPI displays and different OS DPI scaling settings | Use layout managers and size policies; use logicalDpiX() for DPI-aware sizing |
| Setting styles inline on individual widgets | Creates visual inconsistency; extremely difficult to theme or maintain | Define a single QSS stylesheet at QApplication level and use object names/classes |
| Ignoring cross-platform rendering differences | Widget sizes, fonts, and margins differ significantly between Windows/macOS/Linux | Test on all target platforms; use platform-conditional logic where rendering diverges |
# model.py -- Business logic, no Qt dependencies
class DataModel:
def __init__(self):
self._items = []
def add_item(self, item: str) -> bool:
if item and item not in self._items:
self._items.append(item)
return True
return False
# controller.py -- Mediates between Model and View
from PyQt6.QtCore import QObject, pyqtSignal
class Controller(QObject):
items_changed = pyqtSignal(list)
error_occurred = pyqtSignal(str)
def __init__(self, model: DataModel):
super().__init__()
self._model = model
def add_item(self, item: str) -> None:
if self._model.add_item(item):
self.items_changed.emit(self._model._items.copy())
else:
self.error_occurred.emit(f"Could not add: {item}")
# view.py -- UI only, connects via signals/slots
from PyQt6.QtWidgets import QMainWindow, QVBoxLayout, QWidget, QLineEdit, QPushButton, QListWidget
class MainView(QMainWindow):
def __init__(self, controller: Controller):
super().__init__()
self._controller = controller
# Wire signals to slots
self._controller.items_changed.connect(self._on_items_changed)
self._controller.error_occurred.connect(self._on_error)
# UI emits to controller -- never calls model directly
self._add_btn.clicked.connect(lambda: self._controller.add_item(self._input.text()))
def _on_items_changed(self, items: list) -> None:
self._list.clear()
self._list.addItems(items)
from PyQt6.QtCore import QThread, pyqtSignal
class WorkerThread(QThread):
progress = pyqtSignal(int)
finished_with_result = pyqtSignal(object)
error = pyqtSignal(str)
def __init__(self, task_fn, parent=None):
super().__init__(parent)
self._task_fn = task_fn
def run(self):
try:
result = self._task_fn(self.progress.emit)
self.finished_with_result.emit(result)
except Exception as e:
self.error.emit(str(e))
# Apply at QApplication level
app = QApplication(sys.argv)
app.setStyleSheet(Path("styles/dark-theme.qss").read_text())
# QSS file
"""
QMainWindow {
background-color: #2b2b2b;
color: #e0e0e0;
}
QPushButton {
background-color: #3c3f41;
border: 1px solid #555;
border-radius: 4px;
padding: 6px 16px;
color: #e0e0e0;
}
QPushButton:hover {
background-color: #4c5052;
}
"""
# Use layout managers -- never setGeometry() or move()
layout = QVBoxLayout()
layout.addWidget(self._toolbar)
layout.addWidget(self._content, stretch=1) # stretch fills available space
layout.addWidget(self._status_bar)
# For responsive grids
grid = QGridLayout()
grid.addWidget(label, 0, 0)
grid.addWidget(input_field, 0, 1)
grid.setColumnStretch(1, 1) # input stretches, label stays fixed
| Skill | Relationship |
|---|---|
modern-python | Project setup with uv, ruff, ty, pytest |
python-backend-expert | Backend service patterns for desktop app backends |
tdd | Test-driven development for Qt widget testing |
accessibility | Accessibility audit patterns applicable to desktop apps |
Before starting:
Read .claude/context/memory/learnings.md for prior PyQt6 patterns and platform-specific workarounds.
After completing: Record any platform-specific rendering issues, signal/slot patterns, or QThread gotchas to .claude/context/memory/learnings.md.
ASSUME INTERRUPTION: Your context may reset. If it's not in memory, it didn't happen.
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Ensure accessibility in UI components including semantic HTML, ARIA attributes, keyboard navigation, and WCAG 2.2 AA compliance.
The English summary is being prepared. Showing the original description.
Use when you want to improve response quality through meta-cognitive reasoning. Applies 15+ reasoning methods to reconsider and refine initial outputs.
The English summary is being prepared. Showing the original description.
N-round opposing-stance debates for trade-off analysis. Assigns pro/con roles to agents, runs structured debate rounds with quality scoring, and produces a moderator synthesis with confidence-rated recommendation. Generalizable to architecture, technology, security, and design decisions.
The English summary is being prepared. Showing the original description.
Force adversarial code review stance that eliminates confirmation bias — reviewer must find issues or re-analyze
The English summary is being prepared. Showing the original description.
Creates specialized AI agents on-demand when no existing agent matches a request. Use when the Router cannot find a suitable agent for a task. Enables self-evolution by generating persistent agents.
The English summary is being prepared. Showing the original description.
LLM-as-judge evaluation framework with 5-dimension rubric (accuracy, groundedness, coherence, completeness, helpfulness) for scoring AI-generated content quality with weighted composite scores and evidence citations
The English summary is being prepared. Showing the original description.