screen-reader-testing

screen-reader-testing is an agent skill from the wshobson/agents collection that gives a practical guide to testing web applications with screen readers, the assistive technology that many blind and low-vision people use. It covers the major screen readers, including VoiceOver, NVDA and JAWS, because each announces the same page slightly differently and a page can work in one and fail in another. The skill covers validating screen reader compatibility, testing ARIA implementations, where incorrect roles or labels can make an interface confusing, verifying that forms are labeled and their errors are announced, and testing whether dynamic content changes are announced when they happen, such as a status message or a new search result. It also covers checking navigation, including how users move between headings, landmarks and links, and debugging the assistive technology issues that appear in production. The skill applies when validating screen reader support, testing ARIA, debugging assistive technology problems, checking form accessibility, testing dynamic content announcements, and ensuring navigation is usable without sight. It is most useful as a test plan that a team runs with real screen readers, because automated checks miss many of these problems. The agent can write the test checklist and help fix the markup, and people who use these tools should review the results. The skill is installed with the skills CLI from the wshobson/agents repository. Its skills.sh page showed about 10.4K installs and passing results from three security audits at the time of the scan.…