Turns any URL or PDF into clean Markdown without touching local files. It routes intelligently: Feishu and WeChat get special handling, GitHub tries raw content first, Twitter and X go through proxies that preserve image URLs. Everything runs through a proxy cascade to punch through paywalls and JavaScript. By default it just shows you the content inline, but it'll save to Downloads with proper frontmatter if you ask or if you're calling it from /learn. The hard rule is no summarizing, no analysis, just conversion. It also knows when it hit a paywall instead of real content and will tell you rather than save a login page. Useful when you need to quote something, feed content into another tool, or archive without the cruft.
npx skills add https://github.com/tw93/waza --skill read