Skills on AI 459 skills

Active theme: Light

Screen Reader

Use when asked about how screen readers work and how to design/build for them — audio representation of content, semantic markup and alt text, navigation mechanisms, structural/context cues — as the underlying assistive technology screen-reader-testing verifies against.

A screen reader is assistive technology software that helps individuals with visual impairments or blindness access and interact with digital content — a crucial tool for inclusive UI/UX design. See Screen Reader Testing for verifying a specific interface against actual screen readers; this skill covers the technology and what designing/building for it requires.

How it works: an audio representation

Screen readers provide an audio representation of a digital interface's visual content — reading aloud text, icons, buttons, and other elements to enable navigation and understanding without relying on sight.

Designing for screen reader compatibility

Screen-reader compatibility depends on following accessibility guidelines: proper semantic HTML markup (see HTML), alternative text for images, and full keyboard accessibility — a screen reader can only convey what the underlying markup actually expresses, so a visually correct but semantically empty interface (all <div>s, no real headings or landmarks) is effectively invisible to it.

Navigation mechanisms

Screen readers offer users various ways to explore and interact with a digital interface — jumping between headings, links, buttons, and form controls, or navigating linearly, via spoken commands or keyboard input — which is why a logical heading structure and correctly labeled interactive elements matter well beyond just reading order.

Content and context

Screen readers convey the structure and layout of content — heading hierarchy, lists, and other structural elements — letting users navigate more efficiently by structure rather than only by reading everything top to bottom. They also provide context cues: the presence of images, form input requirements, and error messages, which is exactly what ARIA attributes (see Aria Attributes) exist to make explicit for dynamic or non-native-semantic elements.

Common pitfalls

  • Missing or unhelpful alt text — an image with no alt attribute, or one filled with unhelpful text like "image1.jpg," gives a screen reader user no useful information about content a sighted user gets for free.
  • Non-semantic markup — building an interface entirely from generic <div>/<span> elements with visual styling but no real headings, landmarks, or native interactive elements leaves a screen reader with nothing meaningful to announce.
  • Keyboard-inaccessible interactions — a control operable only by mouse/touch (hover-only menus, click-only custom widgets) is unreachable for many screen reader users, who navigate primarily by keyboard.
  • Designing and never testing with an actual screen reader — see Screen Reader Testing; assumptions about what a screen reader will announce, based on the markup alone, are frequently wrong in practice.

Learn more

View screen-reader/SKILL.md on GitHub