feat(blog): why am I not getting interviews (EN + ES), with minimal blog i18n by santifer · Pull Request #119 · career-ops-hq/career-ops-docs

santifer · 2026-09-27T15:40:18Z

What

A new post, Why am I not getting interviews?, in English and Spanish, and the small blog infrastructure the Spanish version needs.

/blog/why-am-i-not-getting-interviews (EN)
/es/blog/why-am-i-not-getting-interviews (ES, same slug)

The post

Opens with a self-contained answer (~60 words) and an anchored index.
US hiring backdrop from the Cleveland Fed Economic Commentary 2026-17 and the BLS July 2026 JOLTS release. Quoted verbatim, dated, labeled as US-only. In Spanish the quotes stay in English with a translation after them.
What the reader controls: evaluate before applying (below 4.0 it recommends against applying), contacto to find one person to write to (it never sends), and the funnel report, follow-ups and patterns (thresholds 20, 5 and three, as in the core).
The author's funnel, signed in the first person, framed as one search and not a benchmark.
It never promises interviews or replies, and never applies in the reader's name.

Blog i18n (minimal)

Same slug under /es/blog, with no /es/blog index while few posts have a Spanish twin.
BLOG_ES_TWINS lists the posts with a Spanish twin. The build fails if that list disagrees with the files, if a Spanish post has no English twin, or if a French post appears.
Reciprocal hreflang (en / es / x-default) only when a twin exists. Other English posts are unchanged.
The Spanish page has inLanguage: es on BlogPosting and FAQPage, Content-Language: es, and LangSync.
Sitemap has both entries with alternates. llms.txt lists the English post, and llms-full.txt stays English-only.

Checked locally (production build)

npm run build passes.
faq-from-mdx --check, verify-agent-layer and verify-schema-parity pass (115 claims on 122 URLs).
EN posts return 200. /es/blog and /es/blog/<post without twin> return 404.
Every index anchor resolves in both languages.

Depends on

Links to /docs/introduction/guides/is-this-job-worth-applying-to (#117). Merge after #117 so those links are not 404 in production. In this preview they 404.

Not to be merged before the 24 h notice to search-ops has run.

🤖 Generated with Claude Code

https://claude.ai/code/session_01BcPLgMFMs2B4SJv3vfq8jv

A post has a Spanish twin iff content/blog/<slug>.es.mdx exists; it is served at /es/blog/<slug> (no /es/blog index), with reciprocal hreflang and x-default=EN, its own canonical, and a sitemap entry dated from its frontmatter. Same loader-only i18n as docs. EN surfaces (index, sitemap, llms-full) now ask for getPages('en') explicitly. The build fails if a Spanish post has no English twin, if a French post appears (no route), or if the language bar's twin list disagrees with the files. The post render moves into a shared BlogPostView; English output is unchanged. IndexNow: date-only lastmods (midnight UTC) are compared by day, or a new post dated from frontmatter would never be announced. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BcPLgMFMs2B4SJv3vfq8jv

A post for people whose applications go unanswered. It opens with a self-contained answer, gives the US hiring backdrop from a Cleveland Fed commentary and the BLS July 2026 JOLTS release (quoted, dated, US-only), and moves to what the reader controls: whether a job deserves an application, who to write to before applying, and what their own funnel shows. It never promises interviews or replies. The Spanish twin is the first post on the minimal blog i18n: same slug under /es/blog, reciprocal hreflang, inLanguage es, FAQPage mirrored from the visible FAQ. The essay links stay on the English essay and say so. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BcPLgMFMs2B4SJv3vfq8jv

vercel · 2026-09-27T15:40:24Z

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
career-ops-docs	Ready	Preview	Oct 9, 2026 7:30am UTC

coderabbitai · 2026-09-27T15:40:30Z

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 12 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Repository: career-ops-hq/career-ops-docs/.coderabbit.yaml
Review profile: ASSERTIVE
Plan: Advanced
Run ID: 9c386dc0-d7bb-4dd8-928e-7722b943a73d

📥 Commits

Reviewing files that changed from the base of the PR and between f217ca5 and db9af31.

📒 Files selected for processing (16)

.github/workflows/indexnow.yml
content/blog/why-am-i-not-getting-interviews.es.mdx
content/blog/why-am-i-not-getting-interviews.mdx
source.config.ts
src/app/(en)/blog/[slug]/page.tsx
src/app/(en)/blog/page.tsx
src/app/es/blog/[slug]/page.tsx
src/app/es/blog/layout.tsx
src/app/llms-full.txt/route.ts
src/app/sitemap.ts
src/components/blog-post-view.tsx
src/components/language-bar.tsx
src/lib/blog-source.ts
src/lib/blog-twins.ts
src/lib/llms-index.ts
src/lib/schema.ts

Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands.}

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BcPLgMFMs2B4SJv3vfq8jv

…llow it On /de (and /es, /fr) the ⌘K dialog searched the English index. /api/search already keeps one Orama index per language, with that language's tokenizer: /api/search?query=bewerbung&locale=de returns 100 German hits, the same query without locale returns 0. But the dialog takes the locale from Fumadocs' I18nProvider, and the shared root layout never mounted one, so it sent none and the server fell back to English. LocaleRootProvider reads the locale from the path, the same way FooterLocaleLink and the AI search do, and passes it to RootProvider together with the ten Fumadocs UI labels for es, fr and de (search placeholder, "On this page", previous/next page...). It passes no `locales` list on purpose: with more than one, Fumadocs layouts render their own language picker next to ours. English is unchanged: locale=en, default labels. No date source is touched (neither file feeds git-date or home-date), so no lastmod moves. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…locale

…ared home modules into _home (renames only) Pure renames, no content change, so every moved file keeps its git history and, with git-date's --follow --diff-filter=r, its lastmod. Edits come in the following commits. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…e server HTML; one home dictionary per language <html lang> was "en" on every page: the single root layout hardcoded it, and a script corrected it in the browser, which crawlers that read raw HTML never see (search-ops D1). Each locale now has its own root layout ((en)/, es/, fr/, de/), all rendering the same RootShell (providers, footer, analytics) with their own lang, the content-language meta on translated pages, and a static locale for the ⌘K search. LangSync and the inline script are gone. Unmatched URLs get global-not-found.tsx (experimental.globalNotFound), which keeps the 404 the site had: Next's message inside the English shell, noindex. Switching language is now a full page load, the cost of separate root layouts. The home dictionary is split: home-dict.tsx keeps the type and helpers, and each language has its own home-<locale>.tsx. A home's lastmod now depends on its route, the shared trunk and its own dictionary, so editing the French home stops moving the English one. Dates: the English routes moved into (en) in the previous commit as pure renames, and git-date now runs git log with --follow --diff-filter=r, so a moved page keeps its last real change as lastmod. The sitemap paths and home-date follow the new locations. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…locale

With one root layout per locale the server HTML carries the right <html lang> on every page, so a wrong one is a regression. Agreed with search-ops when the guard landed (#142): flip it in the same PR as the root layouts. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…locale # Conflicts: # src/components/locale-root-provider.tsx # src/components/root-shell.tsx

Moving the home under (en)/ changes the file name Next derives for its opengraph image (same bytes, new path hash), so the old URL would 404 for cards cached by social networks and for saved links. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…king after the move into (en) An adversarial review of this branch reproduced three regressions: - Every signatory's wide card moved from /manifesto/s/<user>/opengraph-image to ...-1xn4ck (Next hashes generated image routes inside a route group). The example card on /manifesto and /es/manifesto, the "Download your card: wide" link and every card already shared on social sites would 404. A 308 keeps the stable path working, and the i18n guard now fails if either old card URL stops ending in an image (stable-image, with its selftest sabotage). - git-date skipped any rename commit, including R<100 ones. PRs land squashed, so a move and an edit share one commit and the edit never reached the lastmod. -M100% makes only exact renames count as moves. - .coderabbit.yaml still pointed its homepage rules at src/app/(home)/, which no longer exists, so its "visual change needs the maintainer" checks went silent. They now cover src/app/**/(home)/ and src/app/_home/. The outputFileTracingIncludes keys are also written as the escaped picomatch route globs Next documents. The TTF is traced by static analysis either way. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…he site, and drop founderOf from the Person node Bylines, the /about descriptions, the home and /methodology author answers, the press boilerplate, llms.txt and llms-full.txt now use the same descriptor as santifer.io and Wikidata. The author paragraphs are rewritten so they do not repeat it, in en, es, fr and de. /about: Featured in WIRED Greece in the two descriptions, two sentences reworded, Last updated and ProfilePage dateModified set to 9 Oct 2026. /methodology: bio FAQ wording and role-name correction only; methodology unchanged, dateModified kept. schema: founderOf removed from the Person node. It is not a schema.org property, and Organization.founder already states the relation. The signatory bio on the manifesto pages is not touched. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

…he blog i18n branch The blog post page keeps the shared BlogPostView. Its byline now reads creator of career-ops (creador de career-ops in Spanish), and post dates in the sitemap go through handDate. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

santifer · 2026-10-09T07:11:08Z

Status on 9 October.

This PR is now stacked on #149 (the per-language root layouts). #149 moves the blog under src/app/(en)/ and changes the byline, so the two conflicts this branch had with main are resolved against it here:

The post page keeps the shared BlogPostView. Its byline reads "creator of career-ops", and "creador de career-ops" on the Spanish post.
Post dates in the sitemap go through handDate, like the rest of the sitemap.

Built and served locally: 165 of 165 URLs pass the i18n SEO guard (the two new ones are this post in English and Spanish), and the agent-layer guard, schema parity and the future-dates check pass.

Still to do before merging, on the day itself: the post's date and lastModified say 1 October and must be set to the real publication day. Planned merge: Wednesday 14 October after 19:00 UTC, by Santiago, once #149 is in.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

santifer and others added 2 commits September 27, 2026 17:14

vercel Bot deployed to Preview September 27, 2026 15:40 View deployment

santifer and others added 2 commits September 27, 2026 17:41

fix(blog): Spanish opener within the 40-60 word self-contained answer

45c0ebf

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BcPLgMFMs2B4SJv3vfq8jv

fix(blog): Spanish opener within the 40-60 word self-contained answer

c5550e0

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01BcPLgMFMs2B4SJv3vfq8jv

vercel Bot deployed to Preview September 27, 2026 15:42 View deployment

santifer and others added 13 commits October 1, 2026 09:40

Merge remote-tracking branch 'origin/main' into i18n/root-layout-per-…

8446cb4

…locale

Merge remote-tracking branch 'origin/main' into i18n/root-layout-per-…

fb26b64

…locale

Merge remote-tracking branch 'origin/main' into i18n/root-layout-per-…

522ab52

…locale # Conflicts: # src/components/locale-root-provider.tsx # src/components/root-shell.tsx

Merge main (docs fixes of 9 Oct)

a751d64

Merge main (scanner sources fix)

7fadc8c

santifer changed the base branch from main to i18n/root-layout-per-locale October 9, 2026 07:10

Merge main (per-language root layouts are in)

e828811

santifer changed the base branch from i18n/root-layout-per-locale to main October 9, 2026 07:22

vercel Bot had a problem deploying to Preview October 9, 2026 07:22 Failure

santifer closed this Oct 9, 2026

santifer reopened this Oct 9, 2026

fix(blog): remove a leftover merge marker from the post view

db9af31

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

vercel Bot deployed to Preview October 9, 2026 07:30 View deployment

Navigation Menu

Repository navigation

feat(blog): why am I not getting interviews (EN + ES), with minimal blog i18n - #119

feat(blog): why am I not getting interviews (EN + ES), with minimal blog i18n#119
santifer wants to merge 19 commits into
mainfrom
d1-blog-i18n

santifer commented Sep 27, 2026

Uh oh!

vercel Bot commented Sep 27, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented Sep 27, 2026 •

edited

Loading

Review limit reached

Uh oh!

santifer commented Oct 9, 2026

Uh oh!

This branch was successfully deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

santifer commented Sep 27, 2026

What

The post

Blog i18n (minimal)

Checked locally (production build)

Depends on

Uh oh!

vercel Bot commented Sep 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

coderabbitai Bot commented Sep 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review limit reached

Uh oh!

santifer commented Oct 9, 2026

Uh oh!

This branch was successfully deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

vercel Bot commented Sep 27, 2026 •

edited

Loading

coderabbitai Bot commented Sep 27, 2026 •

edited

Loading