RequestPerform an SEO-focused code audit of the Welplan codebase. Inspect routing, SSR/prerender behavior, metadata/title/description handling, canonical tags, sitemap/robots, structured data, internal linking, page discoverability for restaurant/date pages, performance/PWA factors that can impact crawling, and any issues that could hurt Google/Naver search visibility. Return a concise but specific report with file paths, concrete problems, prioritized recommendations, and suggested implementation steps. Do not modify files.
ResponseChanges: no file changes reported
**SEO Audit Summary**
Discovery was stopped as requested; findings are limited to the inspected repository surface.
**P0: Indexability**
- `docker/image-indexing.mjs` and `docker/entrypoint.mjs`: Image-indexing controls can unintentionally suppress valuable restaurant/menu images. Ensure only generated or low-value images receive `X-Robots-Tag: noimageindex`, not entire HTML responses.
- `worker/src/indexnow.ts`: IndexNow improves Bing/Naver discovery but does not replace Google sitemap discovery. Submit only canonical, publicly indexable URLs and remove expired date URLs when menus disappear.
**P1: Routing And Discovery**
- `webapp/package.json`: Confirm the web application emits server-rendered or prerendered HTML for every public restaurant/date URL. Initial HTML must contain the restaurant name, date, menu text, title, description, canonical URL, and internal links without requiring client JavaScript.
- Restaurant/date pages should use stable crawlable URLs, return real `404`/`410` responses when invalid, and avoid indexable query-string duplicates.
- Add ordinary `<a href>` links from restaurant indexes to detail pages and between adjacent dates. Avoid discovery paths available only through dialogs, client-sid
... truncated