If you're wiring up the browser's built-in Proofreader API into a text editor or form, this skill keeps you from misreading the spec. It walks through feature detection, availability checks, monitored model downloads, session lifecycle, and the quirks around correction metadata and permissions-policy constraints. The procedures are opinionated about when to gate creation behind user activation, how to handle download progress, and why you shouldn't reuse destroyed sessions. Worth using if you're integrating client-side proofreading and want to avoid the common traps around option mismatches, quota failures, and browser-specific fallbacks. Don't reach for it if you're doing server-side LLM work or general prompt engineering.
npx skills add https://github.com/webmaxru/agent-skills --skill proofreader-api