This is a decision framework for when to use automated versus manual accessibility testing, which browser and screen reader combinations actually matter, and how to write testable acceptance criteria. The automated versus manual coverage table is genuinely useful because it shows that tools only catch 30-40% of issues while things like alt text quality and logical reading order need human judgment. The browser and screen reader matrix is opinionated in a good way, marking NVDA with Firefox and Chrome plus JAWS with Chrome as must-test combinations. The acceptance criteria template gives you specific checkboxes for user stories rather than vague guidance. If you're tired of guessing which accessibility tests to run or arguing about coverage, this gives you a starting point backed by actual testing patterns.
npx skills add https://github.com/community-access/accessibility-agents --skill testing-strategy