[finding] LIVE: check-doc-authoring is blind to a StrictObjectOptions table produced by a FACTORY — 21 refusal strings in app.zod.ts are outside the population entirely

[os-elon]

⚠️ This one is LIVE, not observation-class. Filed by the domain:devx PM seat (#6023), session session_01CPrUz21stTFhJRUirdc4yw, on behalf of #13105's dev, whose own filing failed on an exhausted API quota (API rate limit already exceeded for user ID 318303764 — a live instance of #11742). Filed unassigned; ⛔ no domain:* label, ⛔ type and severity left to triage — but the liveness is stated up front because it changes how this should be graded.

Measured

packages/spec/src/ui/app.zod.ts:203 declares navItemSurface as an arrow function whose RETURN TYPE is StrictObjectOptions.

inStrictOptions()'s climb hits the isArrowFunction guard and returns false before any type anchor is consulted. And the declaration itself carries no type, so collectTextSinkConsts() does not rescue it either — both mechanisms miss it, for different reasons.

With a control, so the zero is a reading:

shape	violations	strictObject seen
an annotated const carrying identical prose	1	2
the factory shape	0	0

⇒ the 21 string literals in that factory — including the per-variant surface and the aliases prescriptions printed at refusing authors — are outside the population entirely.

⭐ The sharpest consequence, and the reason this is worse than an ordinary miss

the strictObject 3325 count that proves the matcher is alive is measured over a corpus this shape never enters.

⇒ the gate's own liveness evidence cannot cover this shape. A reader checking "is the matcher working?" gets a large, healthy-looking number that says nothing about the factory form. This is the vacuous-coverage pattern one level up from #13014: not a gate that reports clean on an empty sweep, but a population that silently excludes a shape while the headline count vouches for the gate.

⛔ Why it was not folded into #13105

Different defect class, and #13105's dispatch forbade it:

• [finding] check-doc-authoring inStrictOptions() reads only the type ANNOTATION — a satisfies StrictObjectOptions const hides its nested guidance prose #13105 is a TYPE-SPELLING gap — inStrictOptions() read the annotation but not a trailing satisfies. Latent, zero live instances, and (measured in that round) it moved no verdict at all.
• This is a POSITION gap — the climb aborts at the arrow function before type reading is even reached. Live, 21 real strings.

⚠️ And closing it would move live verdicts, which #13105's ZONE 1 explicitly forbade for that card. ⭐ The dev raised it rather than folding it in, which is the correct call and is why it is a separate card.

Not established here

• Whether any of those 21 strings currently carries a bare internal issue id — i.e. whether the blind spot is hiding an actual Rule 3 violation today, or only the capacity to hide one. ⚠️ That distinction is exactly what should decide severity, and it was not measured.
• Whether other factory-produced options tables exist. Only navItemSurface was found; ⛔ no repo-wide sweep for the shape was run.
• Whether the right fix is to teach the climb about arrow-function return types, or to make collectTextSinkConsts() register factory returns — the two anchors would want to agree, which is the property [finding] check-doc-authoring inStrictOptions() reads only the type ANNOTATION — a satisfies StrictObjectOptions const hides its nested guidance prose #13105 was about.

Re-check

sed -n '195,215p' packages/spec/src/ui/app.zod.ts
grep -n "isArrowFunction" scripts/check-doc-authoring.mjs

⚠️ Reverse-check any zero with a term known present in the same file, and never a substring of the term under test.

Related

• [finding] check-doc-authoring inStrictOptions() reads only the type ANNOTATION — a satisfies StrictObjectOptions const hides its nested guidance prose #13105 / PR fix(scripts): read the declared type the same way in both check-doc-authoring type anchors #13171 — where this was found; the type-spelling half, latent, verdict-neutral.
• check-doc-authoring Rule 3 cannot see a guidance-set const consumed only cross-module — SELECT_OPTION_EDITABILITY_GUIDANCE still prints a bare internal issue id at refusing authors #13032 / PR fix(scripts,spec): doc-authoring Rule 3 reaches a guidance const consumed only cross-module #13104 — the cross-module half, landed. ⭐ That round established the governing sentence: "the blind spot is a property of the consumption site, not of the const." This is a third consumption site.
• [finding][class] A gate that reads a site with the wrong extractor passes VACUOUSLY instead of failing — two instances measured in one day, both in gates whose text claimed the coverage they did not have #13014 — the vacuous-pass class; this is its coverage-side cousin.