Dash Dog Chaos Run
Overview:
Dash Dog: Chaos Run is a concept for a fast-paced 2.5D platformer inspired by classic mascot games like Crash Bandicoot, reimagined with a playful dog character and modern design thinking.
The goal was to create a game experience that feels fast, chaotic, and highly responsive, while exploring how AI tools can support both design and development workflows.
The game was prototyped using Claude AI and deployed via GitHub, demonstrating a full concept-to-build workflow.
My Role:
Product Designer & AI-Assisted Developer
- Defined the core gameplay concept and player experience
- Designed mechanics, controls, and interaction systems
- Used Claude AI to generate and iterate on game logic and structure
- Directed gameplay behaviour, physics, and interaction patterns through prompting
- Structured level design, progression, and difficulty scaling
- Tested and refined gameplay for responsiveness and feel
- Deployed and hosted the project via GitHub
Tools:
- Claude AI
- GitHub
- JavaScript / HTML (Canvas)
- Figma
The Problem:
Many platformer games struggle to balance:
- Speed with control
- Simplicity with depth
- Challenge with accessibility
Without careful design, this leads to:
- Frustrating gameplay
- Poor responsiveness
- Low engagement and replayability
The Challenge:
- Designing responsive, satisfying movement mechanics
- Translating gameplay ideas into working code using AI
- Maintaining control over behaviour while using AI-generated outputs
- Balancing chaotic gameplay with clarity and precision
- Creating a playable experience quickly without over engineering
Approach:
- Defined the core gameplay loop (run → avoid → collect → attack)
- Used Claude AI to generate game structure, mechanics, and logic
- Iterated through prompting to refine physics, controls, and interactions
- Tested gameplay frequently to ensure responsiveness and clarity
- Used GitHub to manage, version, and deploy the playable build
- Focused on rapid iteration and learning through building
Solution:
Dash Dog: Chaos Run is a fast-paced platformer where players control an energetic dog navigating obstacle-filled environments.
Key features:
- Fast, responsive movement
Momentum-based running with tight controls - Signature ability: Spin Bark
A spinning attack used to break objects and defeat enemies - Collectibles
Bones, treats, and high-risk rare items - Dynamic obstacles
Moving platforms, traps, and enemies - Level progression
Themed worlds with increasing difficulty - Power-ups
Speed boosts, double jump, invincibility, and magnets
Outcome:
- Successfully created a playable game using AI-assisted development
- Demonstrated ability to translate design concepts into working prototypes
- Reduced development time through AI-assisted coding and iteration
- Showcased a full workflow from concept → build → deployment (via GitHub)
This project highlights how AI can be used not just for ideation, but for building and shipping interactive products.
My Design Process:
- Defining gameplay goals and player experience
- Using Claude AI to generate and refine game logic
- Iterating through prompting to improve mechanics and responsiveness
- Testing gameplay and adjusting based on feel
- Managing and deploying code via GitHub
- Continuous iteration and refinement
Product:
A playable platformer game built using JavaScript and AI-assisted development, demonstrating:
- Fast, engaging gameplay
- Clear interaction design
- Scalable structure for levels and mechanics
- End-to-end creation from concept to deployment
Outcome:
This project demonstrates my ability to combine product thinking, design, and AI-assisted development to create and ship interactive experiences quickly and effectively.
Visuals:
Gameplay screens and interactions are shown below, highlighting mechanics, movement, and level design.
