From Prototype to Pull Request
Jacquelyn Halpern · 18 November 2026
Read the originalSalesforce designer Jacquelyn Halpern says designers at Trailhead now open pull requests that merge into production, and that companies must agree how those changes are reviewed and owned.
Jacquelyn Halpern works on product design and AI at Trailhead, Salesforce's learning platform. Writing in UX Magazine, she describes designers who build prototypes from the product's own components and submit them as pull requests. She sets out what a company must provide before it expects designers to contribute code to the product.
- At Salesforce, designers build prototypes with the code and components of its design systems, and some of their pull requests are reviewed and merged into production.
- An AI agent needs the design system in a form it can use: clear documentation, code examples, consistent component interfaces and guidance on when to use each pattern.
- Code built from production components can work and still not be ready. It must handle missing data, missing permissions, failed requests and keyboard use, and fit the team's architecture.
- Designers need access to the right code repositories, a way to run and preview the product, and representative test data.
A design leader who expects designers to contribute code can agree with engineering what belongs in a pull request, what must be tested, who reviews it and who owns the change afterwards. Halpern gives no figures for how many such changes have merged, and warns that designers can generate more work than a team can review.
Document the design system for AI agents with code examples, consistent component interfaces and guidance on when to use each pattern.
Agree with engineering what belongs in a designer's pull request, what must be tested, who reviews it and who owns the change once it ships.
Derived by Working Surface from the article; more in the Playbook. Source line: What belongs in a pull request? What should we test? Who reviews it? Who owns the change once it ships?
19 November 2026: as AI agents write most of a team's code, the human check moves from reading each pull request to owning the rules on which changes a person reads.