Overview
AI-assisted development patterns and tooling for consistent code generation
The codebase implements AI-first development through context engineering, pattern templates, and automated validation. AI assistants understand the complete architecture and generate code that follows established conventions.
Core Components
The AI workflow relies on four interconnected systems that work together to ensure consistent, production-ready code generation.
Context Documentation
Architecture, security, and development rules in /llms/ guide AI behavior
Pattern Templates
Reference implementations ensure consistent feature development
IDE Integration
Pre-configured files enable immediate AI assistance
Validation Pipeline
Automated checks catch errors in AI-generated code
Context Documentation
Documentation files in /llms/ provide comprehensive codebase knowledge. AI assistants read these files to understand architectural decisions, security requirements, and development patterns.
Pattern Templates
Reference implementations in /llms/templates/ provide complete, production-ready code patterns. Templates include security measures, validation schemas, and error handling.
Pro Tip
AI assistants reference these templates when generating new code. Instead of inventing patterns, they copy and adapt existing implementations, ensuring consistency and security.
IDE Integration
Pre-configured files for multiple IDEs enable immediate AI assistance without manual setup.
Validation Pipeline
Automated checks catch errors in AI-generated code before they reach production. The pipeline runs TypeScript compilation, linting, formatting, and tests.
Heads up
AI-generated code must pass all validation checks before committing. The pipeline ensures type safety, code quality, and test coverage requirements are met.
Development Workflow
The development process follows a structured four-phase approach that ensures quality and consistency.
Research Phase
AI analyzes existing patterns in the codebase. It searches for similar implementations, identifies relevant templates, and understands the architectural context.
Planning Phase
AI proposes an implementation approach based on research findings. The plan references specific files, templates, and patterns to follow.
Plans include file structure, dependency ordering, validation steps, and rollback procedures.
Implementation Phase
AI generates code using identified templates and patterns. Generated code follows existing conventions, includes proper validation, and maintains security requirements.
Implementation occurs task-by-task, with validation checkpoints between steps.
Validation Phase
Automated quality checks verify the implementation. TypeScript compilation, linting, and tests run to catch errors.
If validation fails, AI fixes issues and re-runs checks until all pass.
IDE Setup Guides
Detailed configuration instructions for each supported IDE.
Cursor Setup
Pre-configured .cursor/rules for AI-assisted development
Windsurf Setup
Pre-configured .windsurfrules integration
Claude Code Setup
First-party support with .claude/commands
VS Code Setup
GitHub Copilot integration patterns
Codex Setup
Codex extension and web UI configuration