Visent

Visent is a visual intent layer for AI coding agents. Instead of trying to explain where something is on a webpage or what UI element you want changed, Visent lets you show your coding agent what you mean. Press Alt+S to freeze the current view, draw around the UI you want to change, and describe your intent in plain language. Visent combines your selections with relevant visual, DOM, CSS, and layout context into a structured Visual Intent Package you can paste into your coding agent. HOW IT WORKS 1. Press Alt+S to activate Visent 2. Draw around the UI you want to change 3. Describe the change in plain language 4. Add more selections if needed 5. Click "Copy for AI" 6. Paste the generated context into your coding agent Visent also creates an annotated reference image so your agent can see the exact areas you selected. BUILT FOR AI-ASSISTED DEVELOPMENT Visent is designed for workflows with coding agents such as Claude Code, Codex, Cursor, and similar tools. It is especially useful when you know what you want visually but don't know the exact component name, CSS selector, DOM structure, or frontend terminology needed to explain it. FEATURES • Visual UI selection directly on webpages • Natural-language annotations • Multiple numbered selections • Optional visual reference grid • Relevant DOM and CSS context • Annotated reference PNG • Structured Visual Intent Package • One-click Copy for AI workflow • Ctrl+Z selection undo • Alt+S keyboard activation LOCAL-FIRST Visent has no backend, user accounts, analytics, telemetry, or advertising. Screenshots and UI context are processed locally in your browser. Visent does not send your screenshots or page content to Visent servers. You decide where to paste the generated Visual Intent Package. Point. Draw. Tell your agent what you mean. Like Visent? Visent is an independent project. If it saves you time, you can support its development on Buy Me a Coffee. https://buymeacoffee.com/visent