Signifier
The visible or audible clue that says where an action is and how to use it, like an underline, button shape, label, or drag grip.
See it
What it is
A signifier is the perceivable clue that tells you where an action is and how to perform it. An underline signals a link, a dotted grip signals dragging, and a raised surface with a verb signals a button. Don Norman added the word in the 2013 revised edition of The Design of Everyday Things to separate the cue from the affordance: the affordance is what the object allows; the signifier communicates it. A flat door plate is the signifier that says push, and the word PULL screwed above it is the signifier arguing back.
Use signifiers anywhere the action is not obvious from the object alone. Labels, icons, borders, handles, cursor changes, and partial reveals can work together, but the resting state must carry enough of the message for touchscreens and keyboards, where hover does not help.
Gotcha: the cue makes a promise. Underlined text that is not a link and a card with button styling that does nothing are false signifiers. Icon-only controls can be just as slippery when the symbol is not a widely learned convention. Pair ambiguous icons with visible labels and test whether people can find the action before they use it once.
Ask AI for it
Add clear signifiers to every interactive element in this interface. Render actions with native button and anchor elements, underline links inside prose with CSS text-decoration, give buttons a visible surface and verb label, and add a six-dot grip icon plus cursor: grab to draggable rows. Preserve visible :focus-visible outlines and distinct :hover and :active states, but make every action recognizable before hover. Give icon-only buttons an accessible name with aria-label and add visible text whenever the icon is not a standard convention.