Tenor
Log in

Updates

From a visual reference to working React

A visual reference is most useful when the agent can read its hierarchy, explain what it sees, and turn that understanding into editable interface code.

By

Andres2 min read

The reference is direction

When someone adds an image to Tenor, the goal is not to trace pixels. The image communicates hierarchy, proportion, density, color, and interaction intent. Tenor uses that context to create a component that can still behave like software.

This distinction matters. A screenshot can be visually accurate and still be a poor implementation. A useful result needs real structure, responsive behavior, and code that another agent can understand.

Keep the result editable

The generated component stays on the canvas so the user can keep directing it with plain language. The same artifact can then move into Codex or Claude Code for implementation in an existing project.