How to write a DESIGN.md file Claude can actually useUX Collective
DesignProduct
Guide for creating design documentation that AI assistants can effectively parse.
- AI-readable docs: Your DESIGN.md should tell a cohesive story of your product for AI consumption.
- Structured format: Proper formatting helps AI assistants like Claude understand and work with design decisions.
- Cross-functional tool: Well-structured design docs can bridge communication gaps between teams and AI tools.
For design
Audit your design system documentation format to ensure AI tools can parse and reference it effectively during development handoffs.
UX Collective
DesignProduct
AI products struggle with traditional undo patterns due to non-deterministic outputs.
- Design challenge: Traditional undo functionality breaks down when AI generates different results each time.
- User expectation: Users expect to revert changes, but AI's non-deterministic nature makes this complex.
- New patterns needed: Product teams need to rethink interaction patterns for AI-generated content.
For design
Establish design patterns for version control and state management in AI features before they ship — don't retrofit undo later.
Chrome extension captures webpages and pastes them directly into Figma for editing.
- Workflow streamline: Capture entire webpages or sections and paste directly into Figma with Ctrl+V.
- Design handoff: Eliminates screenshots and manual recreation of existing web elements.
- Immediate editing: Captured elements become editable Figma objects rather than static images.
For design
This type of web-to-design tool could accelerate competitive analysis and design iteration — worth evaluating for team adoption.
Frictionless design colonizes daily life, but artists question what gets lost.
- Optimization critique: Artists and theorists are questioning what's lost as frictionlessness dominates design.
- Aesthetic impact: The pursuit of optimization is reshaping art, aesthetics, and human behavior.
- Counter-movement: Designers are beginning to ask whether some friction might be valuable.
For design
Consider where intentional friction in your products might improve user outcomes or prevent harmful behaviors, rather than optimizing everything away.
Autonomous systems need core moral values, not just logical rules.
- Beyond logic: Autonomous systems require fundamental moral values, not just programmed rules.
- Ethical foundation: Pure logic isn't sufficient for AI systems making decisions that affect humans.
- Values framework: Teams need to establish core principles that guide AI behavior in edge cases.
For product
Define your AI product's core values explicitly during requirements gathering — edge cases will test these values when logic rules fall short.