
Build a design-system prompt library
You'll have a reusable set of AI prompts that embed your team's actual design patterns, so every AI-generated prototype matches your brand instead of producing generic output.
What you'll make
Design System Prompt Library
- Component Prompts: tested prompts for buttons, forms, layouts, etc.
- Pattern Prompts: prompts for common UI patterns and workflows
- Override Rules: where AI consistently gets your system wrong
- Update Protocol: how to maintain prompts when your system evolves
Write each prompt with its specific use case, expected output, and any manual corrections you consistently need to make.
The 3 parts
Each part ends with a check you can see.
Audit Your AI Design Output
Generate prototypes with your current AI workflow and identify where output deviates from your design system.
Done when
You have 5-8 examples of AI-generated components with specific deviations from your design system documented for each.
Build Component-Specific Prompts
Create detailed prompts for your most-used components that include exact specifications and examples.
Done when
You have 5-8 component prompts that each generate output requiring minimal manual correction when tested on fresh use cases.
Test and Document Edge Cases
Use your prompt library on real prototyping tasks and build a documented override list for systematic failures.
Done when
You have used the library on 3-5 actual prototyping tasks with documented override rules for recurring failures and an update protocol for when your design system changes.
Capstone: you put together your Design System Prompt Library.
Where this helps
An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.
About this lesson
What you'll be able to do
- Build a working prompt library with 8-12 tested prompts that generate prototypes matching your actual design system
- Test each prompt against your real components and document what works vs what needs manual override
- Have a repeatable workflow for updating prompts when your design system changes
- Know which design decisions AI can handle vs which require human judgment
Who it's for
- Frontend developers or designers who prototype with AI but get frustrated with generic output
- Design system maintainers who want AI to respect existing patterns
- Teams that have design guidelines but struggle to get AI tools to follow them consistently
What it won't cover
- Building the design system itself — assumes you already have components and guidelines
- Advanced prompt engineering theory or optimization techniques
- Setting up design system infrastructure or tooling beyond prompts
- Multi-model comparison or cost optimization strategies
Skills you'll practise
- Component-aware prompt construction
- Design constraint specification
- AI output validation against existing systems
- Prompt library maintenance workflows