Overview

Creates comprehensive AI coding instructions for the CometVisu project to help AI agents understand the codebase architecture, development workflows, and project-specific conventions.

What's Included

The new .github/copilot-instructions.md documents:

Architecture

• Qooxdoo Framework integration: OOP class system with qx.Class.define(), property system, and event bus patterns
• Dual UI structures:
  • Pure Structure (classic DOM-based widgets with XML parsing)
  • Tile Structure (modern web components with custom elements)
• Backend communication: Multi-backend support (KNX/knxd, OpenHAB, MQTT) with centralized state management via cv.data.Model
• Configuration system: XML-based configs with schema validation and templating

Development Workflows

• Build commands (npx qx compile --watch, npm run source)
• Testing patterns (Karma/Jasmine unit tests, Protractor E2E tests)
• Containerized development environment setup (distrobox)
• Debugging techniques and test mode configuration

Code Conventions

• Qooxdoo class structure patterns with namespacing
• Widget implementation process (class + parser for Pure, custom elements for Tile)
• Naming conventions (PascalCase classes, camelCase properties, __private members, M mixin prefix)
• Required GPL-3.0 license headers
• Asset declaration patterns for external resources

Common Patterns

• Singleton-based state management with address subscriptions
• Plugin system with lazy loading
• Mixin-based architecture (MVisibility, MRefresh, etc.)
• Resource path handling via Qooxdoo ResourceManager

Why This Helps

AI coding agents working on CometVisu will now understand:

• The "big picture" of how the two UI structures differ and when to use each
• Critical build/test commands that aren't obvious from file inspection
• Qooxdoo-specific patterns that differ from standard JavaScript practices
• How the backend communication and state management flows work
• Project-specific conventions like the mixin naming scheme and parser registration

This should significantly reduce the learning curve for AI agents contributing to the project and help them make more appropriate architectural decisions.

Created from VS Code via the GitHub Pull Request extension.

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.