Cursor UI design with real references
Cursor already sees your codebase. Niblet shows it everyone else’s best screens.
Cursor composes from the code it can see. When the task is a surface your codebase has no good example of — an onboarding flow, a paywall, an empty state — Niblet is the missing context: real screens from real apps, retrieved by what they do.
Set up
Connect the remote MCP server
Add the Niblet MCP endpoint to the agent’s user-level MCP configuration with the key itself, not an environment variable. Your dashboard writes this line with your key in place. Four tools are exposed: find_ui_references, find_ui_materials, get_design_reference, and get_ui_component.
# Claude Code claude mcp add --transport http --scope user niblet https://niblet-api.pymodel.com/mcp \ --header "Authorization: Bearer YOUR_NIBLET_KEY"Or install the free local skill
The niblet skill is free, local, and works without an account or token. It carries the reference workflow; pointing it at the MCP adds the live catalogue.
Reference before generate
In Cursor’s agent mode, ask it to pull references first: “find niblet references for this pattern, then propose the layout.” One habit, dramatically less generic output.
The workflow
- Frame
- Name the surface, its job, and the platform you are targeting.
- Reference
- Pull screens with find_ui_references; inspect the full-resolution versions.
- Implement
- Let Cursor map the reference structure onto your components and tokens.
- Review
- Compare the result against the references — hierarchy first, styling second.
Token setup and MCP client configuration are in the docs.
Read the docsQuestions? Contact the project. Legal: Terms · Privacy