docs(service): state how many names the cases views section used to list

[claude]

Fixes #1542

Description

Under Standard list views on the cases page, one paragraph opened three different ways. The English stated six phantom names; both Chinese faces stated a list of seven of which six are phantom. This raises the English to the denominator the Chinese already carried.

The two Chinese faces were right, and they are deliberately left untouched.

What settled it

The card said that deciding this needs the page's own history rather than the current tree. That was correct, and the history answers it precisely — but only the half of the diff that is easy to skip.

Both sentences entered in a single change (3087f211, landed as 11f62429, 2026-08-06). That same change removed the list the two sentences point back at, and the removed side is the referent of "used to list" / 「原先列的」. Reading only the added side is what makes the Chinese look wrong.

The pre-image at 834870cd:content/docs/service/cases.mdx, under ## Standard list views, listed exactly seven names:

Name it listed	A real view?
My Open Cases	Yes — src/views/case.view.ts:126 (my_open_cases), and it is still a row in the views table just above this paragraph
Critical Cases	No
Cases Due Today	No
Breached SLA	No
Recently Closed	No
By Account	No
Service Board	No

Six of those seven map 1:1 onto the six bullets the paragraph still carries, with none left over, and 834870cd:content/docs/service/cases.zh-Hans.mdx carried the same seven. So 「本节原先列的七个名字里，有六个根本不是视图」 is exactly and literally correct. The English was terser rather than wrong: it stated the six and dropped the denominator.

Parity is therefore reached by raising the English, not by trimming the Chinese. Deleting the "seven" would have removed a measured true fact to tidy an arithmetic — the mirror image of the other tempting non-fix, adding a seventh phantom bullet to reach seven, which would have invented a false claim on a page whose whole purpose is naming what does not exist.

The number is durable in a way a roster count is not: it describes what an earlier revision of this page listed, not what the app ships, so it cannot go stale as views are added or removed.

Type of Change

• Documentation update

Related Issues

Fixes #1542

Changes Made

• content/docs/service/cases.mdx:183 — one line. It now names the seven-name denominator and My Open Cases as the one that was real and survived, then leads into the same six unchanged bullets.
• .changeset/cases-phantom-count-states-the-denominator.md — added, carrying the measurement for the release-notes reader.
• content/docs/service/cases.zh-Hans.mdx and .zh-Hant.mdx — not edited. They were already correct.

Testing

• Unit tests pass — pnpm verify exit 0, the whole chain
• Linting passes — included in that chain
• Build succeeds — included in that chain
• Manual testing completed — not applicable, prose-only diff
• New tests added — see the note below

pnpm verify (validate && typecheck && lint && lint:i18n-gate && hygiene && hygiene:tokens && build && test) ran green on 16de80dc, each stage printing its own verdict:

• ✓ Validation passed
• ✓ i18n lint gate: 0 i18n/missing-* issues
• ✓ source hygiene clean (including ✓ no raw control bytes in first-party files)
• ✓ source token ratchet clean
• ✓ Build complete
• Test Files 160 passed (160) · Tests 3391 passed | 1 skipped

⚠️ None of that is evidence that the count is now right. No gate in this repository reads this paragraph's prose, and no test was added because the scope for this card is documentation only. What backs the number is the hand count recorded above, cited to ref and path: the seven-name pre-image at 834870cd, the 1:1 mapping onto the six bullets on 16de80dc, and my_open_cases in src/views/case.view.ts.

One measurement caveat, stated rather than hidden: the clone is shallow. Its single boundary is 814dc372, which is neither 11f62429 nor its parent 834870cd; 834870cd resolves as a commit and 834870cd:content/docs/service/cases.mdx resolves as a blob and reads with real content, so the pre-image is a direct positive read and needs no control leg.

Additional Notes

Re-measured before pushing: origin/main is still 8223d0aa, so this branch sits one commit ahead of the tip with no rebase needed.

While measuring this paragraph I also measured the sibling page that the parallel card owns: content/docs/service/sla-and-escalation.mdx:144, and both of its Chinese faces, say crm_case ships seven views and then enumerate seven, omitting Unassigned — triage (src/views/case.view.ts:213, unassigned_triage). The views table on this page says eight and lists it, so that sibling page carries a stale roster count in all three faces. That page is out of scope here and was not touched — #1541 remains open and owns it.

---

Generated by Claude Code

---

Generated by Claude Code

[vercel]

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

1 Skipped Deployment

Project	Deployment	Actions	Updated
hotcrm	Ignored		Sep 5, 2026 11:38pm UTC