Skip to content

Latest commit

 

History

History
57 lines (42 loc) · 1.89 KB

File metadata and controls

57 lines (42 loc) · 1.89 KB

Roadrunner Autocoder Documentation

📋 Project Overview

Roadrunner is a comprehensive autocoding platform that transforms natural language descriptions into functional code through AI-powered planning, brainstorming, and execution workflows.

🏗️ Architecture

Following AGENTS.md principles with strict modular design:

Input → Process → Output
Prompt → Validate → Result  
Question → Explore → Apply

📁 Module Structure

Core Modules

  • /core - Central orchestration and engine logic
  • /services - Backend API integration layer
  • /stores - Reactive state management
  • /composables - Reusable logic components

Feature Modules

  • /components/planning - Visual workflow builder
  • /components/brainstorming - Multi-agent ideation
  • /components/execution - Code generation and editing
  • /components/configuration - Settings and customization

Support Modules

  • /utils - Pure utility functions
  • /assets - Icons, images, fonts
  • /styles - Design system implementation

🎯 Design Principles

  1. No Inline Code - Logic separated from presentation
  2. No Inline CSS - Styled components with Tailwind
  3. Always Modularize - Feature → Component → Function
  4. Never Assume Completion - Full validation and fallbacks
  5. Always Reference Code - Comprehensive documentation
  6. Comment Every Section - Purpose and rationale explained
  7. Maintain Visual Neatness - Consistent structure and naming
  8. Embrace Meticulousness - OCD-grade attention to detail

🚀 Getting Started

See individual module READMEs for specific setup instructions.

📊 Success Metrics

  • Modularity: Each feature as independent module
  • Configurability: All settings externally configurable
  • Iconification: Consistent icon system throughout
  • Performance: Sub-second response times
  • Usability: Intuitive workflow for all user types