Databutton Launches Point and Prompt for Instant App Editing.

Databutton introduced Point and Prompt, a visual editing feature that lets users click any element in their app, describe what they want changed, and see updates apply instantly.

The AI-powered vibe coding platform announced on LinkedIn that the feature cuts iteration time by 70% based on early testing. The workflow is straightforward: point to a button, form field, or other UI component, type what should change, and Databutton's Milo AI applies the modification without requiring manual code edits.

The feature targets a pain point familiar to no-code and AI-assisted development: translating visual changes into working code. Traditional approaches require either describing the entire context in a prompt or manually editing generated code. Point and Prompt creates a direct connection between what users see and what they want to change.

Users click an element to select it, which gives Milo AI the necessary context about that specific component, its current state, and its relationships to other parts of the application. They then describe the desired change in natural language. The AI interprets the instruction within that narrow scope and updates the element accordingly.

The 70% reduction in iteration time suggests that pointing eliminates much of the back-and-forth typically required to specify which element needs modification and where it lives in the component hierarchy. Rather than writing "change the submit button on the contact form to blue," users can simply click the button and write "make this blue."

Databutton has steadily expanded how users interact with its AI agent. The company launched Oneshot for rapid app prototyping and has been quick to integrate new AI models as they become available. Point and Prompt represents a shift toward more granular, contextual control.

While platforms like Lovable, Bolt, and v0 rely heavily on text-based prompting for UI generation, Databutton is betting that combining visual selection with natural language will reduce ambiguity. The approach mirrors how designers work in tools like Figma, where selecting an object before modifying it is standard practice.

The feature arrives as AI coding tools increasingly focus on reducing friction in the edit cycle. Tools like Cursor and Windsurf use inline suggestions and multi-file editing to speed up development; Point and Prompt extends similar logic to visual builders.

Point and Prompt is live in Databutton's web interface at databutton.com. The company did not specify whether the feature works with all component types or has limitations on the complexity of changes it can handle in a single prompt. Early access users have been testing it, according to the announcement, though wider user feedback is not yet public.

Databutton positions the feature as the fastest way to fix issues in apps built with Milo AI, suggesting it is designed primarily for iterative refinement rather than initial construction. The company continues to operate its Riff product for enterprise agent development alongside the consumer-facing Databutton platform.