Yezur Wiki:Daily routine/Report

Working report of the Alompan daily routine (how it works). Every finding lives as one row below until it is resolved: set a row's Status to give a verdict, and the next run carries it out and clears the row. Nothing on the wiki is changed except what is explicitly approved or amended.

The possible statuses:

• pending — awaiting review; no action is taken (the default)
• approved or accepted — carry out the Suggested fix, as written
• amended — carry out the fix described in Notes instead
• deferred — real, but not now; moves to Dismissed flags and is not re-flagged
• rejected — not an issue; moves to Dismissed flags and is not re-flagged
• to to-do — real issue that should be marked on User talk:Yezur's to-do list instead for human intervention
• fixed — real issue, but was a simple fix for Yezur, or has by other means already been taken care of; confirm and clear

To revive a dismissed flag, move it back into the table as pending.

Flagged items

#	Flagged	Type	Severity	Page(s)	Issue	Suggested fix	Status	Notes
72	2026-08-04	structural	medium	270 Dictionary: entries (Proto-Unterian)	Every Proto-Unterian entry states where it came from in prose, and not one of them says so to Lexemes. The 270 entries written this window each carry an ===Etymology=== section naming their source — 257 of the form "From Proto-Andusian {{m|YCP-XX|…}}" and 13 of the form "A Proto-Andusian-age formation, X + Y, continued only in Unterian". None passes etym_type or etym_source to its {{head}} call, so all 270 rows reach the Lexemes table with both fields empty. Yezur Wiki:Conventions gives that table one job — "cognate sets, borrowing chains, … one query each" — and the query it exists for, every Proto-Unterian word inherited from Proto-Andusian, currently returns nothing, on a language where the answer is 270. Ancient Hertic is the comparison: 209 of its 212 rows carry both fields, and the three that do not carry {{etym-missing}} instead and sit correctly in Category:Ancient Hertic entries needing etymology — that language files its filiation, this one does not. The other four languages with entries are root protos or single lemmas, so Proto-Unterian is the wiki's only descendant language whose descent is unrecorded in structured data.	Mechanical, and asserts nothing the entries do not already say. Add |etym_type=inherited|etym_source=YCP-XX to the {{head}} call of the 257 entries whose etymology reads "From Proto-Andusian …", and |etym_type=compound|etym_source=YCP-XX to the 13 formations — the same two values already in use in the table (inherited 209, compound 1). The source is read from each page's own etymology line, not inferred, and no definition, gloss or heading changes. One pass; the routine can do it the day it is approved.	pending	etym_type=inherited|etym_source=YAP-XX, read off each page's own etymology line. The other 276 Gaillean entries are correctly empty: they carry {{etym-missing}} and belong to the review queue of #122, not to this row. Ancient Hertic remains the comparison and remains the only language that files its filiation — 209 of 212. Update 2026-08-19: Lexemes now holds 3,239 rows and this row's scope 1,367, and both movements are small and fully accounted for. Gaillean's rows fell 504 → 501 when #120 and #121 were carried out this morning, but its contribution here is unchanged at 226: the eleven head calls removed were duplicates, and every one of the 226 Gaillean pages that names a Proto-Herta-Olaic etymon in prose still names it and still files it in no field. Besoh gained one lemma — Dictionary:soǵob, written under #89 — taking its blank lemma rows from 332 to 333. The four Andusian languages stand at 270, 269, 269 and 333. Ancient Hertic remains the only language on the wiki that files its filiation, 209 of 212.
73	2026-08-04	structural	low	Yezur Wiki:Maps	The one page in the project namespace that is not a project page. Asked to check whether every Yezur Wiki: page belongs there (see #47, dismissed), the routine read all six. Five are policy or project documentation and sit correctly: Yezur Wiki:Conventions, Yezur Wiki:Daily routine and its Report, Yezur Wiki:Dictionary structure, and Yezur Wiki:Main Page. The sixth, Yezur Wiki:Maps, is a software smoke-test for the YezurMaps extension: it opens "YezurMaps extension test" and "If you can read this but see no maps below, the module didn't attach — check Special:Version lists 'YezurMaps'", then runs four {{#map:}} calls to prove the parser function fires, that both tile backends are reachable, and that Overpass geometry loads. It addresses the reader directly throughout ("Replace 42 with a REAL relation ID from your data", "a route you've actually mapped") and its content is placeholder test data — First test province, Test marker with popup, relation ids 42 and 148–150. It is also the only page in the namespace that is uncategorised and the only one that nothing on the wiki links to.	flag-only. Where a live extension test belongs is a decision the convention set reserves to the operator, and it is the concrete case #47 was really about: if Meta: is to hold anything, this page is the readiest candidate, and moving it would give that namespace its first page and the sidebar's Project docs link somewhere to lead. Three answers are all defensible — move it to Meta:, leave it where it is and categorise it, or delete it now that the extension is known to work. The routine holds the move right and can carry out whichever is chosen.	pending	New 2026-08-04, and the direct product of #47(amended), which asked exactly this question. Recorded as not a finding: Yezur Wiki:Main Page is uncategorised too, but that is #14 (deferred), which covers the portal front pages as a class.
74	2026-08-04	naming	low	Template:Infobox polity, Yezur Wiki:Conventions	The house word was settled this run, and the one template that uses the other word is the one nothing uses. Carrying out #51 recorded the operator's ruling that country is the house word in preference to polity, and rewrote both worked examples on Yezur Wiki:Conventions to read "Countries in Andusia". Template:Infobox polity is now the only place on the wiki still carrying the retired word — and it is one of the three entries on Special:UnusedTemplates, transcluded by no page at all. Category:Countries holds one member (Engil) and already uses the settled word; Template:Infobox former country and Category:Former countries already agree with it too. The Polities Cargo table the template once declared no longer exists — it was removed with the rest of the encyclopedia's Cargo (see #43, dismissed) — so no data anywhere is keyed to the name.	Cheap, and cheapest now: rename Template:Infobox polity to Template:Infobox country, leaving a redirect, and update the one word in its documentation. Nothing transcludes it, so nothing can break, and the routine holds the move right. The alternative is equally clean: say that polity is wanted as a distinct term — a living state as against a country in the broader sense — in which case the template is correctly named and the row closes with a sentence on Yezur Wiki:Conventions saying so. What is not wanted is the present state, where the convention page names one word and the only template on the subject names the other.	pending	New 2026-08-04, and the residue of an operator ruling rather than a fresh defect. The ruling settled the worked examples, which is what #51 asked; this is the consequence it left behind. #43 (rejected) is not re-opened — that row was about the Cargo table, which is gone; this is about the name on the template.
75	2026-08-04	structural	low	Herta-Olaic, Njeshan, Tacchoan, List of languages	Two hand-maintained lists now write the same twelve names two different ways, because two verdicts said so. #67 was approved as written, and its fix says the missing Members rows go in "as plain text where no article exists … so no red link is created". #71 was amended in the same triage, and its note says the opposite for the roster — "red-link the rest: link all twelve roster entries". Both were carried out this run, exactly as instructed. The result is that Airananwe, Anilonic, Senitian, Usinger, Esselitian, High Gaillean, Kankai, Middle Airananwe, Old Ousperian, Old Usinger, Skennic and Terjiquan are blue-linked on List of languages and plain text in the Members table of the family article that lists them. Every other name in those same tables is linked. The two lists agree perfectly on membership — the check confirms 6/6, 8/8 and 7/7 against the roster — and disagree only on this.	Mechanical either way, and one pass. (a) Link them in the Members tables too, in the [[X language|X]] form the roster now uses — matching every other row of those tables and making the two lists identical in form as well as content. This adds no new red-link targets: all twelve are already wanted from the roster. (b) Or unlink them on the roster, if the red links there were meant to be temporary. (a) is the reading that follows from #71, which is the later verdict.	pending	New 2026-08-04. Filed rather than assumed: both instructions were explicit and scoped to different pages, so the routine carried out each as given rather than harmonising them on its own. Not the same as #57 (deferred), which is about whether the four Tacchoan branch names get articles — these twelve are languages, and the roster already treats them as wanted.
76	2026-08-04	factual	low	Dictionary:kidz, Dictionary:kizi, Dictionary:oʊmjiəɲɔ	Three entries call the peoples of Yezur human, and nothing on the wiki says whether that is the word. The three are the only in-frame uses of human anywhere on the wiki, checked across all 1,264 pages: "# person, human being" on the Proto-Andusian entry and on its Proto-Unterian descendant, and "# a human being, someone; one of us as against a beast" on the Proto-Herta-Olaic one. Every other occurrence of the word on the wiki is out-of-frame, describing the human editor. The three are consistent with each other and read naturally; the question is only whether the encyclopedia's own voice uses human for Yezuri people, which touches Yezur Wiki:Conventions §The Earth test — the same section that rules on "Yezuri" as a reflexive qualifier — and would settle how every future entry glossing person is written.	flag-only. A canon question, not a maintenance one: the routine will not choose a word for the peoples of the world. Say whether human is in-frame vocabulary. If it is, the row closes and nothing changes. If it is not, the fix is three definition lines and the routine can make it in one pass.	pending	New 2026-08-04 as a row, though not a new observation: it was recorded inside #62's notes on 2026-07-30, when there was one such entry. That row was cleared this run, so the question would otherwise have gone with it; it is filed here so it has a home of its own. It has also grown from one entry to three, all three by inheritance from the same gloss. Update 2026-08-06: grown again, from 3 pages to 5 — 6 uses in all, still the only in-frame uses of the word on the wiki, now across all 2,419 pages. The additions are Dictionary:kij (Besoh) and the Proto-South-Andusian section of Dictionary:kidz, both inheriting the same gloss from the same Proto-Andusian entry. Every new descendant language will inherit it again. Update 2026-08-07: 7 uses on 5 pages, and the seventh is a new shape. The other six all gloss person as "person, human being", inherited from one Proto-Andusian entry. The new one is on Dictionary:o' — "claw, talon (of beasts) — the human nail is kio'" — where the word is not a gloss at all but does taxonomic work in a definition, dividing the peoples of Yezur from the beasts. That is a larger claim than the other six make, and it arrived in the same edit as #86.
77	2026-08-04	structural	low	Dictionary:About Besoh, Dictionary:About Proto-Andusian, Dictionary:About Proto-North-Andusian, Dictionary:About Proto-South-Andusian, Dictionary:About Proto-Unterian	The Dictionary has grown a class of appendix page and no category for it. Five Dictionary:About … pages now document a language apiece, four of them a proto-language — its conventions, phonology, sound laws from its parent, derivation and reduplication. They are the only Dictionary pages that are not entries, they are written to a common shape, they cross-link one another, and all five are uncategorised: they are five of the eight entries on Special:UncategorizedPages. No category on the wiki fits them — Category:Dictionary maintenance is for review queues and is hidden, and the {{head}}-emitted lemma categories only apply to entries. More will follow: three further proto-languages are registered in Module:Languages and two of them have no About page yet.	One category, then four one-line edits — but the name and placement are the operator's call, as Yezur Wiki:Conventions reserves new categories. The shape is plain: a category for Dictionary pages that describe a language rather than a word, parented to Category:Languages or to a Dictionary branch of it. Once named, the routine can create it and file the four.	pending	New 2026-08-04. Distinct from #14 (deferred), which covers the three portal front pages and the root Category:Yezur — the other three names on the same list. Recorded as not a finding: these four pages are also on Special:LonelyPages and Special:DeadEndPages by the same measure as every Dictionary entry. Update 2026-08-06: a fifth page joined the class, Dictionary:About Besoh, and it is the first that documents an attested language rather than a proto — so the category's name should not assume proto. All five remain uncategorised, and they are five of the eight entries on Special:UncategorizedPages. Update 2026-08-18: seven pages now, not five. Dictionary:About Gaillean and Dictionary:About Middle Airananwe joined the class this window, both written to the same shape as the other five, both cross-linking them, and both uncategorised — so the class is now seven of the ten entries on Special:UncategorizedPages, and the three that remain are the portal front pages of deferred #14. The second of the two documents a language with no entries at all (see #127), which is a further argument that the category wanted here is for pages that describe a language rather than for pages that accompany a body of entries.
78	2026-08-04	structural	low	File:Tacchoan-subfamilies.png, Tacchoan	The file's description is the article's caption, copied, and the caption has just changed. File:Tacchoan-subfamilies.png's description page was the Tacchoan §Members caption word for word. Carrying out #55(a) rewrote that caption this run — it now describes what the map shows and names Herta-Olaic as the neighbouring family rather than a branch — while the description page still carries the old wording, because nothing keeps the two in step. #54 asked three questions about file description pages; the operator answered two of them (files are categorised; a source line where one is needed), and this is the third, left open: whether a description page may restate an article caption or should point at it. The wiki has one file today, so the answer costs one edit; #53 (deferred) means a corrected export is coming, which will make it two.	Mechanical once decided, and the decision is small. (a) Point rather than restate: the description page says what the file is and where it is used, and leaves the colour key to the article. (b) Or restate, and accept that the two are kept in step by hand — in which case the routine can copy the new caption across now and check the pair on every run. Either way the file's categories, added this run, stay as they are.	pending	New 2026-08-04. Not a re-opening of #54, which was carried out as amended and cleared: this is the one question the amendment did not reach, filed so it is not lost. Not a re-opening of #53 either — that row is about the image itself, this is about the page describing it.
79	2026-08-06	markup	medium	Dictionary:About Besoh	One bullet on the new appendix page is broken in the reader's face and wrong about the Dictionary. The §Entry conventions bullet reads: "Nouns carry a Declension section ({{tl|YCQ-BE-decl}}), verbs a Conjugation section ({{tl|YCQ-BE-conj}}). The bare plural of a noun has its own entry (a non-lemma form, without etymology); other table forms do not." Two things are wrong with it. (a) {{tl}} — the template-link helper — does not exist on this wiki. Both calls therefore render as the same red link, Template:Tl, so the sentence reads "Nouns carry a Declension section (Template:Tl), verbs a Conjugation section (Template:Tl)" and names neither table. It is the page's only red link, and the only wanted template outside the four Ancient Hertic paradigms of deferred #21. (b) The claim about entries is no longer true. The Dictionary holds 648 Besoh non-lemma entries in six shapes, not one: 125 Plural of, 130 Dative singular of, 127 Dative plural of, 125 Locative singular of, 68 Locative plural of and 66 Stative of. The bullet was written at 08:18; the first inflected-form entries followed eleven minutes later and the last at 10:17. The page was edited again at 14:31 — the window's final edit — but that edit added the grammar and word-formation sections and left this bullet as it stood.	One edit carries both. (a) Drop the helper and name the tables in plain italics or <code> — which is also what the other four Dictionary:About … pages do, none of which names a template at all. Creating Template:Tl would fix the red link but leave an in-frame page pointing at the wiki's own machinery, which is the deeper half of (a). (b) Replace "The bare plural of a noun has its own entry … ; other table forms do not" with the six shapes actually in use — plural, dative singular, dative plural, locative singular, locative plural, and the verbal stative — or with a general clause ("single-word inflected forms have their own entries; periphrastic forms do not"), which matches Module:Besoh's own linking rule and will not go stale the next time a form is added.	pending	New 2026-08-06. Filed as one row on purpose: (a) and (b) are in the same bullet, and two verdicts scoped to one sentence is exactly how #75 happened. Both halves are mechanical and neither asserts anything the wiki does not already say — the six form shapes were counted from the entries themselves, and the linking rule from the module's own edit summary. Not a criticism of the page otherwise: its five sound laws, ten spoken developments, orthography and grammar all check out against the entries and against Module:Auphen/YCQ-BE. Update 2026-08-07: unchanged, and now measurably more wrong. Dictionary:About Besoh was edited three times this window (07:25, 08:01, 10:05), each time adding grammar, and neither half of this bullet was touched. (a) {{tl}} still does not exist and still collapses both calls to the same red link. (b) the entry claim is now false of 797 non-lemma entries in the same six shapes — 159 plural, 159 dative singular, 159 locative singular, 157 dative plural, 156 locative plural, 85 stative. Update 2026-08-18: half (a) has spread to a second page. Dictionary:About Gaillean, created this window, writes its §Entry conventions bullet as "Nouns carry their class in the headword line and a declension table ({{tl|YAQ-GA-decl}})" — the same non-existent helper, rendering the same way, so the sentence reads "a declension table (Template:Tl)" and names no table. Template:Tl is now transcluded by two pages, both of them Dictionary:About … appendix pages, and it remains the only wanted template outside the four Ancient Hertic paradigms of deferred #21. Whichever way (a) is answered — create the helper, or drop it and name the tables in plain markup — the answer should be applied to both pages in one pass. Half (b) is unaffected and unchanged. Re-checked 2026-08-20: the same call is now live on two further pages, written after this row was filed — see #134, which asks only that whatever verdict this row receives be read as covering them too.
80	2026-08-06	naming	medium	Dictionary:About Besoh	An in-frame page uses the one word the convention set names as forbidden. Yezur Wiki:Conventions §No real-world references says it outright — writing systems are not to be invoked by their real-world names, "no 'Latin script', no 'Romanized'". The first bullet of §Entry conventions uses the word twice: "Entries currently live at romanized titles" and "entries will move to native-script titles with romanized redirects". Checked across all 2,419 pages, these are the only two in-frame uses of it on the wiki. The three others are all out-of-frame and all permitted: Yezur Wiki:Conventions naming the ban, Yezur Wiki:Dictionary structure describing the same practice, and a comment inside Module:Auphen/YCQ-BE. The same bullet also says the move will happen "once its type is prepared for the wiki" — the encyclopedia describing its own production rather than its subject.	(a) Reword, in place. The wiki has no house term for the script it currently writes Yezuri languages in, so the minimal fix avoids naming one: "Entries currently live at transliterated titles" … "entries will move to native-script titles, with the transliterated forms kept as redirects", and "once type for it exists" for the clause about the script. (b) Or settle the house word first, and apply it to both this page and Yezur Wiki:Dictionary structure, which describes the same practice in the same borrowed word — out-of-frame and so permitted, but it is where the term entered. A single word, chosen once, closes both.	pending	New 2026-08-06. The strictest rule in the set, and the only one the convention page enforces by naming the offending word, which is what makes this checkable rather than a matter of taste. Not re-opening #12 (rejected), which settled that Greek-style native names are deliberate in-world orthography: that row is about what the scripts look like, this is about the word used to describe writing one script in another. Update 2026-08-07: unchanged. The page took three edits this window without the bullet changing, and romanized remains the only two in-frame uses of the word on the wiki, now measured across all 2,622 pages.
81	2026-08-06	maintenance	low	50 Dictionary: entries (Besoh plural forms)	Fifty more pages joined the deletion queue, which now stands at 92. A ruling this window merged word-final e and o into the plural suffix -u, so the twenty-five bare-plural entries written earlier in the day at -o'u / -e'u titles, and their twenty-five dative-plural companions, were superseded within hours of being created. Each carries {{delete|Superseded — the plural of X is now Y (pluralization rule change, operator ruling)}}. The replacements all exist — checked one by one, all twenty-five -u entries are present — and nothing links to the old pages: all fifty are on Special:LonelyPages, so no table, entry or list points at a title that is about to disappear. The queue is now: 5 talk pages (#58, deferred), 37 Proto-Andusian entries at implosive titles (#70, deferred), and these 50. The titles, unlinked so this page adds no incoming link to a page queued for removal: bazo'u, bazo'ubi, ce'u, ce'ubi, cuzze'u, cuzze'ubi, debe'u, debe'ubi, gisbo'u, gisbo'ubi, gośme'u, gośme'ubi, ho'u, ho'ubi, hoffo'u, hoffo'ubi, humme'u, humme'ubi, izvo'u, izvo'ubi, konso'u, konso'ubi, kotme'u, kotme'ubi, makvo'u, makvo'ubi, medo'u, medo'ubi, mukko'u, mukko'ubi, raco'u, raco'ubi, ruve'u, ruve'ubi, se'u, se'ubi, sevo'u, sevo'ubi, sorpe'u, sorpe'ubi, tuqme'u, tuqme'ubi, ve'u, ve'ubi, vutme'u, vutme'ubi, zebe'u, zebe'ubi, ñoro'u, ñoro'ubi.	flag-only — only an administrator can act on it. The editing bots hold no delete right, so this row exists to put the fifty in front of the one account that can remove them, and to be dismissed the way #58 and #70 were, so later runs list them under Deletion queue without re-flagging. Nothing else is wanted: the tags are correct, the reasons are stated, the replacements are in place.	pending	New 2026-08-06. Recorded as not a finding beside it: the reasons say "operator ruling", which names the wiki's process inside an in-character namespace — but Template:Delete's own documentation declares the tag "out of the in-universe frame; exempt from the Watsonian rules", so the wording is within its own licence and is not flagged. Also not a finding: all fifty were created and superseded inside the same window, which is churn rather than error — the rule changed after the batch, and the batch was corrected the same day. Update 2026-08-07: the queue is 94, the two additions being Dictionary:kunke and Dictionary:kunkeva, superseded by the kunke → kiunke respelling and correctly tagged. But two of the fifty must not be deleted — see #82. Under the new pluralization rule the titles ho'u and ho'ubi are the live plural and dative plural of Dictionary:ho', a different lemma whose declension table links to them; clearing the queue as this row lists it would red-link that table. Of the 94 pages queued, exactly three have any incoming link at all: those two, and the kunke of #84.
82	2026-08-07	structural	high	Dictionary: entries ho'u and ho'ubi (titles unlinked, see below), Dictionary:ho'	Two of the fifty pages queued for deletion are the live plural and dative plural of a different word. The pluralization ruling of 2026-08-06 moved the plural of Dictionary:ho (ho, “egg”; “stone”) from ho'u to hu, and the two superseded titles were blanked and tagged — correctly, for ho. But Dictionary:ho' is a separate lemma (ho', “guts, entrails”), it ends in a glottal and not a vowel, and Module:Besoh's rule (-u after a consonant) therefore gives it the plural ho'u and the dative plural ho'ubi — the same two titles. Its {{YCQ-BE-decl}} table links to both, and they are the only two incoming links either page has: of the 94 pages in Category:Pages flagged for deletion, exactly these two and one other (#84) are not on Special:LonelyPages. So the two pages now (a) carry a deletion banner in the reader's face where the plural of ho' belongs, (b) state a reason — "the plural of ho is now hu" — that is false of the lemma that actually points at them, and (c) will red-link a live inflection table the moment the queue is cleared. Checked against every Besoh lemma: 159 nouns and 85 verbs, 244 lemmas and 875 inflected forms — these are the only two collisions on the wiki, and apart from them not one live Besoh form is missing a page.	Mechanical, and it should be done before the queue is cleared. Rewrite the two pages as ordinary ho' non-lemma entries in the standard shape — {{head|YCQ-BE|noun form|head=ho'u}} and a definition line "Plural of ho'" / "Dative plural of …", matching the 795 other form entries — which removes the {{delete}} tag as a side effect and leaves nothing for an administrator to do. The old ho sense is already served by hu and hubi, both written and correct, so nothing is lost. What must not happen is the deletion going ahead as #81 lists it: that removes two pages a live table depends on.	pending	New 2026-08-07. The one row this run that blocks another: #81 puts all 50 superseded plurals in front of the administrator, and 48 of them are safe. These two are not, and the reason is a homograph the rule change could not have anticipated — ho and ho' are different words that the old rule sent to the same plural and the new rule separates. Verified rather than reasoned: the rule was read out of Module:Besoh itself, the incoming links from Special:WhatLinksHere, and the replacement forms confirmed present. The two titles are written unlinked here, as in #81, so this page adds no incoming link to a page queued for removal. Update 2026-08-19: the evidence behind this row is now cleaner than when it was filed. Of the 122 pages in the queue, exactly two in the Dictionary: namespace are still absent from Special:LonelyPages — the two this row is about — and the third page it had to name as an exception, the superseded kunke, joined that list this morning when #84 was carried out and its last incoming link was repointed at kiunke. (The five Talk: pages of deferred #58 sit outside the namespaces that page covers.) Nothing else moved: both titles are still blanked to a bare {{delete}} tag, still carry a reason that is false of the lemma whose live declension table links them, and ho' is still the only page linking either. This window touched neither.
86	2026-08-07	structural	low	Dictionary:o', Dictionary:kio'	An entry was narrowed and its structured gloss was not, so two Besoh nouns now file the same meaning and one of them denies it. Dictionary:o' was edited this window to narrow it to beast-claws — its definition now reads "claw, talon (of beasts) — the human nail is {{l|YCQ-BE|kio'}}" (the human in it is the new use counted under #76) — but its {{head}} call still says gloss=fingernail, and that is what reaches Lexemes. Dictionary:kio', the word the definition hands the sense to, files gloss=fingernail as well. The table therefore holds two Besoh nouns glossed fingernail, one of which says in its own definition line that it is not one. Checked wiki-wide, this is the only lemma on the wiki whose head gloss appears in no definition line of its own section and where the two genuinely disagree: the other 25 the check raised are all the Proto-Herta-Olaic and Proto-Njeshan style where the gloss is a keyword and the definition is a description of it (fingernail → "nail, claw; the hard growth at a digit's end"), which is house practice and not drift.	One parameter. Change Dictionary:o''s head to gloss=claw, matching its own first definition line and the word it is now contrasted with. Nothing else on either page changes, and Dictionary:kio' is already right.	pending	New 2026-08-07. A near-miss recorded rather than flagged: Dictionary:go is the same shape — its primary sense moved to "whole, full; the complete count" and its head still says gloss=ten — but ten does survive there as a labelled archaic sense, so the gloss is stale rather than contradicted, and the page has a live verdict pending against it already (#83). If #86 is approved and the operator wants the same done to Dictionary:go, say so in Notes and both go in one pass.
87	2026-08-07	structural	low	12 Dictionary: affix entries (Proto-Andusian)	The Proto-Andusian affixes promise thirty-six branch descendants and not one of them arrives. The twelve affix entries — -dzi, -gɛ, -kɛ, -me, -oŋ, -so, -uɣ, -wa, -ɛts, -ᵐbu, sog-, ᵐbi- — each carry a Descendants list naming their North Andusian, South Andusian and Unterian reflexes. All 36 pointers resolve to nothing, and they fail in two different ways. Twelve are ordinary red links, at eleven distinct titles (-ets, -ge, -mu, -oŋi, -u, -uɣi, -zi, -õ, -ɛsi, so-, mi- — -mu is wanted twice, by both the South Andusian and the Unterian reflex of -ᵐbu), which Special:WantedPages reports and which are 11 of the 60 names on it. The other twenty-four are worse than red: where the reflex is spelled the same as the ancestor (-me, -so, -wa and six more), the link is blue — it lands on the affix page the reader is already on, at a section anchor that does not exist there. Nothing reports those: not Special:WantedPages, not Special:BrokenRedirects, not the linter. The three branch protos are otherwise complete, at 269, 269 and 270 lemmas apiece, so this is the one part of the family's vocabulary the seeding did not reach.	flag-only. Writing 36 affix entries is lexicographic work, not maintenance, and the routine will not invent morphology. Two smaller answers are also open if the entries are not wanted soon: the 24 same-spelling pointers could be plain italics rather than {{desc}} links until the sections exist, which costs one edit per affix page and stops the reader being sent in a circle; or they can stand as a standing wanted-list, which is what they are today.	pending	New 2026-08-07. Not a re-opening of the wanted-pages pattern the last run recorded as self-resolving — that reading was right, and it is why wanted pages fell from 319 to 92 to 60 as the languages were seeded. This row is the remainder after that fall: the affixes are what is left, and the twenty-four blue-but-dead pointers are the half of it no maintenance page can see. The count was taken by resolving every {{desc}} call on the wiki — 1,140 of them — against the Lexemes table.
90	2026-08-08	markup	medium	Dictionary:About Besoh	Two bullets swap bold for italic in the reader's face, and both lose a Besoh letter doing it. An italicised word ending in the glottal apostrophe closes with three apostrophes in a row, and on a line that also carries a bold run the parser pairs them the wrong way round. Two lines of §Participles and subordinate clauses do this. The active-participle bullet is written * '''Active participle''' in ''-so'' (the agent suffix in clausal duty): it stands before its noun and takes its own object to its left — ''korsu muqso bo''' “a man driving horses”. and renders as: Active participle' in bold, with a stray apostrophe; then the whole explanation in italics; then korsu muqso bo in bold with its final glottal eaten, so the example reads bo, a different Besoh word, in place of bo' “man”. The privative bullet does the same: Privative' in bold, the explanation italic, and sagobbe' rendered as sagobbe, which is not a word. §Orthography is a third, milder case — the glottal's own row, written '''' '''' = ʔ, renders as a loose apostrophe and a bold space. A balance check does not catch any of this: the page closes 29 <b> and 142 <i> tags cleanly. The tags are paired; they are simply around the wrong words.	Three one-line edits, and no wording changes. Use the pattern the same page already uses correctly in §Basic grammar for the negator and the perfect particle: put a space before the closing italics, ''korsu muqso bo' '', or write the apostrophe inside <nowiki> tags. Applied to the two bullets and the orthography row, the bold returns to the term being defined, the explanation returns to upright, and both Besoh examples get their glottal back.	pending	New 2026-08-08, and found by rendering rather than by reading: the wikitext looks unremarkable. Swept wiki-wide, the pattern occurs on seven pages outside this Report, and on five of them it renders as intended because no bold shares the line — Dictionary:-ke and Dictionary:-so among them, both checked and correct. It mis-renders on this page, in the two places above, and on User talk:Yezur, which is out-of-frame and not filed. Distinct from #79 — a different bullet on the same page and a different defect — but the two are one sitting.
91	2026-08-09	voice	medium	Ancient Hertic language, Hertia, Ignusia	Three articles call the eastern inland sea "Yezur's cradle of civilization", and the wiki's two other mentions of the same place do not. Yezur Wiki:Conventions sets the Earth test in as many words: "Wherever 'Yezur' or 'Yezuri' appears, the sentence should still read as a natural Wikipedia sentence with 'Earth' in its place … a reflexive 'Yezuri' in front of every other noun does not" pass. An encyclopedia of Earth writes the cradle of civilization, never Earth's, because there is only one and the possessive adds nothing. Three pages take the possessive — Ancient Hertic language ("a region often described as Yezur's cradle of civilization"), Hertia ("remembered as Yezur's cradle of civilization") and Ignusia ("The eastern sea is Yezur's cradle of civilization") — while List of languages writes "in a region often described as the cradle of civilization" and Main Page writes "along the eastern one lies the cradle of civilization". Five mentions of one place, three of them possessive and two not, so the fix is toward practice the wiki already keeps. Everything else passes. Every occurrence of Yezur and Yezuri in the 50 main-namespace pages was read against the test: "a continent in the southwest of Yezur", "the tallest mountains on Yezur", "Yezur's polar continent" and "loanwords drawn from languages across Yezur" all read naturally with Earth in place, and the last of those is the Conventions page's own worked example of a passing sentence. This construction is the only one that fails.	Three one-word edits. Replace "Yezur's cradle of civilization" with "the cradle of civilization" on Ancient Hertic language, Hertia and Ignusia. Nothing else in the three sentences changes, and no claim about the world is altered — the sea is still the cradle of civilization, and it is still the only one.	pending	New 2026-08-09, from the first wiki-wide reading of the encyclopedia against Yezur Wiki:Conventions. Earlier runs surveyed the change window, and every window since this Report began has been Dictionary work, so the 31 articles in the main namespace had never been swept as a set. Recorded and deliberately not flagged: Sihia's Yezuri Age of Exploration and Taccho's Yezuri Middle Ages carry the same reflexive prefix, but they are dismissed #8 — the operator has ruled these are not the canon names for the eras — so the routine leaves them alone rather than tidying a name that is going to change. New World, on Lovvocia and Vetsia, is already an open thread on both Talk pages ("a placeholder, awaiting a Yezuri derivation"), recorded exactly where the Conventions say an open thread belongs, and is not re-raised here.
92	2026-08-09	markup	low	Andusia, Herta-Olaic, Hertia, Njeshan, Shung-Worrehan	Ten boldface runs in the encyclopedia mark something that is not the article's subject. Yezur Wiki:Conventions on text formatting: "Boldface marks the article's subject — and its significant alternative names and redirect targets — at first mention, and almost nothing else. Do not bold terms for emphasis." Every bold run in all 31 main-namespace articles was matched against its page's title and against every redirect pointing at that page; ten fail, on five pages. Seven are branch names bolded where they are introduced — Herta-Olaic (Hertia branch, Olay branch), Njeshan (Betonia branch, northwestern branch, southwestern branch) and Shung-Worrehan (Shung branch, Worrehan branch). Two are bolded wikilinks in Andusia's subcontinent bullets, on Unteria and Betonia, where the bold adds nothing the link does not already do. One is an alternative spelling: Hertia writes [[Phose Serou]] (''Φοσε Ζερου''; also spelled '''Fose Serou'''), putting the second spelling in bold beside the native form in italics. Tacchoan is the comparison: it introduces four branches — Esselis, Rumesselis, Kipoa and Usy — in plain text, so the house form already exists on the wiki and four of five family articles simply disagree with it. And the bolding that is correct is left alone: English language bolds Traditional English and New English, and both are redirects to that page — the one case the convention names as allowed.	Five edits, no wording changed. Take the bold off the seven branch names and off the two subcontinent links, leaving the links themselves in place. On Hertia, move the alternative spelling into italics beside the native form: [[Phose Serou]] (''Φοσε Ζερου''; also spelled ''Fose Serou'').	pending	New 2026-08-09, from the same wiki-wide sweep as #91. Main Page is excluded from the count on purpose: it is a designed portal rather than an article, and its bold is layout, not prose emphasis. The 19 redirects in the namespace hold no prose and were skipped.
93	2026-08-09	structural	low	Besoh language, Lonish language, Nichana language, Pjany language, Uu language	Five of the eleven language articles cannot be reached by the name the rest of the wiki calls them. Yezur Wiki:Conventions on naming: language articles take the "X language" form, and "Native names and other aliases become redirects." Six of the eleven have the bare-name redirect — Ancient Hertic, English, Gaillean, High Nichana, Otjesi and Pises — and five do not: Besoh, Lonish, Nichana, Pjany and Uu are unclaimed titles. The split is not random. Four of the five missing ones are exactly the articles that got a native-name redirect instead — Lôniêv, Nehchánh, Peanei and Uu-Co all resolve, so each of those four received one alias and not the other. Besoh language received neither, and is the only language article on the wiki reachable by no alias at all: its native name Bëson is given in its own lead and in its infobox and has no page, and the bare title has no page. It is also the wiki's most-written language, with 332 lemmas and 797 form entries in the Dictionary.	Create six redirects, each a single line, matching the six that already exist: Besoh, Bëson, Lonish, Nichana, Pjany and Uu — titles written unlinked here so this Report does not put six fresh red links into the wanted-pages count — every one #REDIRECT [[X language]]. No article is edited and no link changes, because every article link already goes through the piped "X language" form.	pending	New 2026-08-09, from the same sweep as #91. None of the five is a red link — nothing on the wiki links a bare language name — so this repairs no broken link and costs no article edit; it is reader-facing navigation and search only, and that is why it is filed low. One judgment the routine will not make for the operator: Nichana and Pjany are plausible names for a future people or region, and Uu is short enough to want for something else. If any of the bare titles should stay free, say which, and the row narrows — Besoh and Bëson have no such competition and are the part of this fix that is unambiguous.
94	2026-08-09	structural	low	Tacchoan, Herta-Olaic	The five language-family articles are written to one shape, and two of them break it in one place each. Four of the five head their sections Branches, Members, Distribution, in that order. Tacchoan has no Branches heading at all: its four branch bullets — Esselis, Rumesselis, Kipoa, Usy, each with a sentence of range — sit under == Members == above the language table, so a reader looking for the branches finds them filed under members, and a reader looking for the members meets a branch list first. Herta-Olaic alone heads its closing section == Colonial distribution == where the other four use == Distribution ==, although all five sections do the same job and three of the other four describe colonial spread in them.	Two edits, no prose touched. On Tacchoan, rename the existing == Members == to == Branches == and insert a fresh == Members == immediately above the language table; the image stays where it is, at the head of the branch section it illustrates. On Herta-Olaic, rename == Colonial distribution == to == Distribution ==.	pending	New 2026-08-09, from the same sweep as #91. The routine takes the four-article majority as the house form rather than deciding which is better writing; if the more specific heading is the wanted one, the amendment is to rename the other four instead, and the routine will carry that out just as readily. Not flagged beside it: Andusian gives its Members section one sentence of prose where the other four use a table — that is dismissed #11 (the family has one attested member so far), and the sentence is accurate.
95	2026-08-09	structural	medium	Main Page	Half the front page's navigation leads nowhere. The Explore the world panel offers four cards of identical design — same border, same rule, same bold heading. Two carry a link: Lands and seas to Category:Continents and Tongues and families to Category:Language families. The other two, Peoples and pasts ("The nations, ages, and deeds of the world") and Life and nature ("The creatures, plants, and wild places of Yezur"), carry bold text and no link of any kind. Nothing in the design tells a reader which two are live, so the first thing a first-time visitor does on the wiki's front page has an even chance of doing nothing.	Half mechanical, half a decision. Peoples and pasts has targets already on the wiki and can be linked today: Category:History (its own description reads "former countries, periods, and events") covers the pasts half and Category:Countries the peoples half — which of the two the card should point at is the operator's pick, and the routine will wire whichever is named. Life and nature has no target: all 37 categories on the wiki were read and none covers living things or wild places. The honest options for it are to leave the card as it stands, to remove it until there is somewhere for it to lead, or to create the category first — and a new category needs the operator's say-so under Yezur Wiki:Conventions, so the routine will not invent one.	pending	New 2026-08-09, from the same sweep as #91, and the only finding in it that a reader meets before anything else. Filed medium for that reason rather than for its size. Recorded and not flagged: Main Page is also uncategorised, which is dismissed #14, covering the portal front pages as a class.
96	2026-08-09	factual	low	Alompe, Hertia	Two sentences read "the Ancient Hertic", with nothing for the article to attach to. Alompe closes "…and the Ancient Hertic was spoken there" and Hertia closes "…and the birthplace of the Ancient Hertic." In both, the word language sits inside the pipe, so what the reader sees is the article's name with its noun taken away. Herta-Olaic shows the pattern that works, twice on one page: "The Ancient Hertic language belongs to the Hertia branch" and "the Gaillean language belongs to one of these" — the noun sits outside the link and the sentence reads. Checked across all 50 main-namespace pages: these two are the only piped "X language" links preceded by a bare the with no noun after them.	Two edits, one word each, both outside the link. On Alompe: "…and the Ancient Hertic language was spoken there." On Hertia: "…and the birthplace of the Ancient Hertic language." The link target, the display text and every fact stay as they are.	pending	New 2026-08-09, from the same sweep as #91. Filed as factual rather than markup because the missing word is in the sentence, not in the wikitext: the markup is well-formed and renders exactly what it was told to render.
97	2026-08-10	structural	low	The 12 pages of the Module: namespace	The wiki's twelve modules are the only body of pages on it that no category holds. Every other namespace is filed: 28 of the 31 templates carry one of four categories (Category:Templates, Category:Dictionary templates, Category:Infobox templates, Category:Link templates), all 34 documentation pages of the Map: namespace sit in Category:Map documentation, the one file carries two, and four of the six project pages carry Category:Yezur Wiki (the two that do not are #73 and dismissed #14). The modules carry none — Module:Auphen, its three sound-data pages and Module:Auphen/frame, Module:Besoh, Module:Head, Module:Infobox, Module:Infobox/doc, Module:Labels, Module:Languages and Module:Translations, twelve for twelve. And nothing on the wiki reports it. Special:UncategorizedPages reads only the content namespaces, which here are the main one, Dictionary: and Map:; the module namespace is outside its reach, so the eight names that page lists have never included one. No category for them exists or is wanted: all 105 entries on Special:WantedCategories are per-language lemma and part-of-speech categories emitted by {{head}} and {{lb}}, and not one names a module or a template.	One category, then twelve one-line edits — but the name and placement are the operator's call, as Yezur Wiki:Conventions reserves new categories. The plain shape is a container for the wiki's Lua modules, sibling to Category:Templates under Category:Yezur; Category:Modules is the obvious name and is encyclopedia furniture of the kind §Naming already permits ("Infobox, Cite, Stub, Navbox"). Once named, the routine can create it and file all twelve in one pass.	pending	New 2026-08-10, from the first reading of the Template:, Module: and Category: namespaces as a set. Not a re-opening of #25, which the operator sent to the to-do list at User talk:Yezur and which the routine does not re-flag: that item is about seven modules whose /doc subpage does not exist, and its fix is writing documentation. This is about the category footer, and its fix is one line per page. The two do meet, and that is worth saying: on a wiki that carries module documentation, the category conventionally sits on the /doc page rather than in the module source, so if the doc pages are written first this row is carried out inside that work and costs nothing extra. If they are not, the twelve lines stand on their own.
98	2026-08-10	structural	low	Template:Translations/testcases	The count of uncategorised templates this Report prints every run is three, and only one of them can be fixed. Special:UncategorizedTemplates has read 3 for as long as this page has recorded it, and the routine has carried the figure without ever resolving it into names. The three are Template:Infobox/styles.css, Template:Translations/styles.css and Template:Translations/testcases. The first two are not defects: both are stylesheets, held in the sanitized-css content model, in which a category link is not wikitext but a syntax error — they cannot be categorised and will sit on that list permanently. The third is ordinary wikitext, and its own sibling shows what it should carry: Template:Auphen/testcases, the wiki's other live test page, ends with [[Category:Templates]].	One line. Add a category to Template:Translations/testcases. Category:Dictionary templates is the closer fit — the template it tests, Template:Translations, is filed there, while Template:Auphen/testcases follows Template:Auphen into Category:Templates — so either answer keeps a test page beside the template it tests. After it, the honest figure for uncategorised templates is 0 of 29 that can be.	pending	New 2026-08-10. Filed apart from #97 although both are furniture categorisation: this one is a single line on a single page needing no decision, and #97 cannot start until a category is named. Recorded as not a finding beside it: Template:Auphen/testcases and Template:Translations/testcases are two of the three names on Special:UnusedTemplates, which is correct — a test page is meant to be transcluded by nothing. The third is Template:Infobox polity, which is #74.
99	2026-08-10	voice	medium	Yezur Wiki:Conventions, Yezur Wiki:Main Page	The frame table does not reach the namespaces the wiki's machinery lives in, and its rule of thumb gives them an answer they cannot keep. Yezur Wiki:Conventions §Namespaces names six: (Main), Dictionary: and Map: in-character, Yezur Wiki:, Meta: and Talk: out-of-frame. Those cover 2,533 of the wiki's 2,622 pages. The remaining 89 are in Category: (37), Template: (31), Module: (12), User: (7), File: (1) and MediaWiki: (1), and the section's closing rule of thumb rules on them by omission — "everything a reader or editor sees is in-character, except the Meta:, Talk:, and Yezur Wiki: namespaces. That includes template, module, and category names" — so all 89 are in-character. Twenty of them are not, and four say so themselves. Six pages name a real-world reference work in as many words: Template:Delete ("Modelled on Wikipedia's deletion tags"), Template:L, Template:M and Template:Lb ("Wiktionary-style", "curated from Wiktionary's own label set"), Module:Labels and Module:Languages — eight occurrences in all. Eleven of the twelve modules are Lua source, and their comments necessarily name their own materials: five uses of Latin, seven of IPA, two each of Greek and ASCII, and in Module:Labels the word Earth itself, in a line explaining that everything Earth-specific was dropped from its label set. And the four that declare their own frame do it without a rule to point at: Category:Dictionary maintenance and Category:Yezur Wiki say in their visible text that they are "out-of-frame — about the wiki, not about the world", Category:Pages flagged for deletion says so in a comment, and Template:Delete goes furthest, its comment reading "out of the in-universe frame; exempt from the Watsonian rules" — an exemption the convention set does not define. The wiki's second statement of the rule disagrees with the first: Yezur Wiki:Main Page §Conventions says "world content in Main and the Dictionary (in character); project talk in Meta, Yezur Wiki, and Help, and on Talk pages (out of character)", which drops Map: from the in-character side altogether and covers none of the 89 either.	flag-only. What frame the machinery namespaces are held to is a convention decision and the routine will not make it — but it is the rule the routine needs, because it decides whether a phrase like Wiktionary-style in template documentation is a finding or house practice, and today the routine has no answer to check against. The shape of the fix is small: three or four rows added to the §Namespaces table. The distinction that already exists in practice is between the name of a template, module or category — which §Naming governs, and which every one of the 89 keeps — and its documentation or description, which is editor-facing and where all twenty exceptions sit. If that is the intended line, saying so in the table closes the row, legitimises the four self-declarations, and lets Template:Delete's comment cite a policy instead of asserting one. The second statement on Yezur Wiki:Main Page should be brought into step with whatever is decided.	pending	New 2026-08-10, from reading the three furniture namespaces end to end for the first time. Not #47 (deferred), which is about Meta: and Help: being documented while holding no page — this is about namespaces that hold 89 pages and are not documented. Not #73 either, which asks where one particular project page belongs. Recorded as clean beside it: the Map: namespace, which the table does declare in-character, was read in full and holds zero real-world references across all 35 pages — the frame is kept everywhere it has been stated.
100	2026-08-10	naming	low	Template:Infobox language, the 11 language articles	Every language article shows a field label its own template calls a placeholder. Template:Infobox language renders a row headed Language code, carrying the registry code from Module:Languages — Language code: YCQ-BE on Besoh language, and the same on all eleven language articles, every one of which fills the code field. The template's documentation says what it is: "The code field holds the language's registry code (label "Language code" is a placeholder, to be renamed once an in-universe term is coined)". So the wiki knows the label is temporary, and the note recording that sits in template documentation, where Yezur Wiki:Conventions §Open threads does not put open threads — they are "recorded as plain text on the article's Talk page, under an == Open threads == heading". Checked across every Talk page on the wiki: none mentions it. The five family articles are unaffected — Template:Infobox language family has no such field.	Two answers, and they are independent. (a) The note: move the open thread to a Talk page, so it sits where the convention set puts one and is not lost when the documentation is next rewritten — Talk:List of languages is the natural home, the roster being the page that tabulates the codes, and it does not exist yet, so the title is written unlinked here; the routine can create it as soon as it is told which page. (b) The label: coining the in-universe term is the operator's, and the row stays open until it is coined. Recorded for whenever that happens: the change is one word in one template and reaches all eleven articles at once, because the label is written once and rendered eleven times.	pending	New 2026-08-10. Filed low, not medium: nothing is broken, nothing contradicts, and a code registry is a thing an encyclopedia may plausibly keep about its own languages. What makes it a row rather than nothing is that the template says the label is provisional and nowhere else on the wiki does. Not #37, which is on the to-do list at User talk:Yezur and asks whether reconstructed languages get articles; that row is about the registry's contents, this is about the label on the field that displays them.
101	2026-08-10	structural	low	Category:Hidden categories	A wiki-wide container is filed inside one branch of the wiki it is meant to span. Category:Hidden categories is the tracking category MediaWiki fills automatically: any category carrying __HIDDENCAT__ lands in it without being told to, which is why Category:Auphen errors and Category:Auphen test failures are members although neither page names it, and why the container is a member of itself. Its own description says as much, and is accurate. Its parent is not: its one category link is Category:Dictionary maintenance, a branch whose own description scopes it to "the Yezur Dictionary and the modules that serve it". Today the three hidden categories on the wiki are all Dictionary ones, so nothing is misfiled; the container is. Hide any category outside the Dictionary — a deletion queue, a file backlog, a maintenance list for the map — and it will appear, automatically, inside a Dictionary branch. Of the 37 categories on the wiki this is the only one whose stated scope is wider than its parent's; the other 36 nest inside theirs, and the Category:Ancient cities → Category:Ancient cities in Hertia chain is the worked example Yezur Wiki:Conventions §Categories describes.	One line. Reparent Category:Hidden categories from Category:Dictionary maintenance to Category:Yezur Wiki — which is Category:Dictionary maintenance's own parent, so the container moves up one step and stops being a Dictionary page while the three maintenance categories it holds stay exactly where they are. Nothing else changes and no page's own categories move.	pending	New 2026-08-10, from mapping all 37 categories and their parents for the first time. Recorded as not a finding beside it: the three members sit in both Category:Hidden categories and its ancestor Category:Dictionary maintenance, which the redundant-categorisation check raises — but one of those two is added by hand and the other by the software, so it is a by-product of the tracking mechanism and not a miscategorisation. Also not a finding: Category:Auphen errors and Category:Auphen test failures are the two names on Special:UnusedCategories and both are empty, which is the good state each of them declares in its own description.
102	2026-08-10	markup	low	Template:Desc	The one line of documentation this template has runs the example instead of showing it. Template:Desc's documentation opens "A Descendants-section line: {{desc|YCP-XC|weki}} renders the language's display name … followed by a link to the word, starred for reconstructed languages." The call is wrapped in <code> and not in <nowiki>, so the parser executes it. What the page actually renders is "A Descendants-section line: Proto-Unterian: *weki renders the language's display name …" — the output stands where the syntax should be, and an editor reading the page is never shown how to call the template. The sentence still reads as English, which is why it has gone unnoticed: the substitution is invisible unless the page is rendered and compared to its source. Every other dictionary and link template gets this right — Template:L, Template:M, Template:D, Template:Head, Template:Lb, Template:Etym-missing, Template:Auphen and the four map-link templates all wrap their examples in <nowiki> or a <pre> block, and several then show the live output deliberately, after an arrow. Template:Desc shows only the output.	One line, no wording changed. Wrap the call: {{desc|YCP-XC|weki}} inside <nowiki> tags, as the sibling templates do. If the live rendering is wanted as well — it is a fair demonstration, and the entry it reaches, Dictionary:weki, exists and is correct — follow the house pattern the other templates use and write the escaped syntax, an arrow, then the live call on the same line. Either way it costs no orphan: the live call is one of that entry's two incoming links, and the other, from Dictionary:bʰek, keeps it off Special:LonelyPages whichever way this row is carried out.	pending	New 2026-08-10, and found by rendering rather than by reading, like #90: the wikitext looks ordinary and the rendered sentence is grammatical, so only holding the two side by side shows the swap. Recorded as not a finding beside it: Template:YCQ-BE-conj and Template:YCQ-BE-decl carry a one-line description with no worked example at all, where their Ancient Hertic counterparts show one — but both are called bare, with no arguments, so there is nothing to exemplify. Template:YAQ-HE-decl/core likewise shows none, and says in its own comment that it is not to be placed on entries directly.
103	2026-08-10	structural	low	Template:Head	The wiki's most-used template reads nine parameters, documents seven, and one of the two it leaves out is in live use on 143 entries. Template:Head files every Dictionary entry into Lexemes — 2,737 rows — and its documentation gives a usage block for 1, 2, gloss, register, etym_type and etym_source, plus a definition list for 1, 2 and head. Two more are read and never mentioned. class is passed to Module:Head and, where the module knows the code and part of speech, prints the gender or inflection class inline and files a category for it — it is used on 143 pages, all Ancient Hertic, in seven values (F 36, OO 28, IS 28, M1 20, N1 14, M2 11, N2 6), and it is the parameter behind category names like Category:Ancient Hertic first-declension masculine nouns, which are among the 105 on Special:WantedCategories. notes is both read by the template and declared in its Lexemes schema, and is used by no page at all. The only description of class anywhere is a comment in the Lua source of Module:Head — and that module is one of the seven with no /doc page, so an editor who opens Template:Head to learn how to write an entry cannot find it. The mirror case is worth naming: register is documented and used by no entry, so the doc and the practice have drifted in both directions.	Two lines added to the documentation, and nothing else touched. Add class to the usage block and the definition list — "inflection class, where the language defines one; shows the class inline and files its category (Ancient Hertic: F, M1, M2, N1, N2 for nouns; IS, OO for verbs)" — read off Module:Head's own table, not invented. Add notes beside it, or, if it is not wanted, drop it from the template and from the Lexemes declaration — but that second choice needs the one-time administrative recreate data rebuild Yezur Wiki:Conventions requires for any schema change, so documenting it is much the cheaper answer.	pending	New 2026-08-10. Not a re-opening of #25, the module-documentation item on the to-do list at User talk:Yezur: that is about writing /doc pages for seven modules, this is two lines on a template that already has documentation. The two connect only in that Module:Head's missing /doc is why the Lua comment is the sole record of class. Recorded as not a finding: the five infobox templates were checked the same way and every one documents exactly the parameters it reads — 7 of 7, 15 of 15, 6 of 6, 5 of 5 and 11 of 11 — so this is the one template on the wiki where the two lists differ.
104	2026-08-11	structural	low	Hertia, Khurou, Thiorosirega, Phose Serou	Three of the four cities in Hertia's list link back to the list. Hertia §Notable cities names four, each with its native spelling: Alompe (Αλουμπε), Khurou (Χουρου), Thiorosirega (Θηοροςιρεγα) and Phose Serou (Φοσε Ζερου). Alompe is an article. The other three are redirects, and all three point at Hertia#Notable cities — the very section the reader clicked from, so the link returns them to the line they started on with nothing to say they have not moved. Nothing on the wiki reports this: a redirect to the page that holds the link is neither broken nor double, and Special:BrokenRedirects and Special:DoubleRedirects both read 0. Every wikilink on all 2,622 pages was resolved through the redirect table this run, and these three are the only self-returning links on the wiki. The redirects themselves are not the problem and should stay — they catch the three names from search and from outside — and all three anchors are live: Notable cities is a real heading on Hertia, as are the anchors of the other two anchored redirects.	Two answers, and the wiki has already chosen one of them once. (a) Leave the three names plain in the bullets, keeping the native forms and the redirects — which is exactly what Tacchoan does with its four branch names under dismissed #57, where the operator's ruling was that the names "stay plain in the bullets that define them" until articles exist. Three lines changed, Alompe's bullet untouched, and the routine can do it in one pass. (b) Or write the three as articles on Alompe's model — a sentence or two, the native name, {{On the map}} and Category:Ancient cities in Hertia — which is lexical and geographic content work, not maintenance, and is the operator's to make. The category is the measure of the gap: it holds one member for the four cities the article names.	pending	New 2026-08-11, from the first link-topology check the routine has run — earlier sweeps resolved links against pages that exist, not against where a redirect lands. Not #12 (rejected), which settled that the Greek-letter native spellings on these very lines are deliberate in-world orthography: that row is about what the names look like, this is about where the links go. Not #92 either, although it touches one of the same bullets — that row takes the bold off Fose Serou — but if both are carried out they are one edit on one line. Recorded beside it: all 20 redirects on the wiki are bare one-liners, carrying no category and no text of their own, so nothing else is riding on these three.
105	2026-08-11	structural	low	Sea of Sihia, Unteria, Betonia	Three articles are named by a placeholder, and the note saying so is on somebody else's Talk page. Yezur Wiki:Conventions §Open threads is exact about where such a note goes: unresolved questions and placeholder names "are recorded as plain text on the article's Talk page, under an == Open threads == heading". Three of them have outgrown that. Talk:Andusia carries "'Unteria' is a placeholder, awaiting a Yezuri derivation" and "'Betonia' is a placeholder, awaiting a Yezuri derivation"; Talk:Sihia and Talk:Vetsihia both carry "'Sea of Sihia' is a placeholder, awaiting a Yezuri derivation". But Unteria, Betonia and Sea of Sihia are each an article in their own right now, and not one of the three has a Talk page — so an editor who opens the article whose title is the placeholder finds nothing, and the thread lives only on the pages that mention it in passing. All 32 main-namespace articles were checked against the 19 pages of the Talk: namespace: 12 articles have a Talk page, and these three are the only articles on the wiki that are the subject of a recorded open thread and have none.	Three Talk pages, one line each, copied from where the thread already sits — an == Open threads == heading and the placeholder bullet in the wording Talk:Andusia and Talk:Sihia already use. Nothing new is asserted, no article is edited, and the routine can do it in one pass. Leave the existing notes where they are: Unteria and Betonia are subcontinents of Andusia and the Sea of Sihia is enclosed by Sihia, so those pages have their own reason to carry the thread. This adds a note; it does not move one.	pending	New 2026-08-11, from the first reading of the Talk: namespace since 2026-07-26, before the article and Dictionary work that created these three pages. Not #59 (rejected), which was about these same three articles reaching no Cargo table and finding no infobox to fit a subcontinent, an island or a sea — that question closed with Cargo, and this one is about a Talk-page note. Consistent with #91's reading rather than against it: that row recorded New World as "an open thread on both Talk pages … recorded exactly where the Conventions say an open thread belongs", and it is — New World is not an article, so Talk:Lovvocia and Talk:Vetsia are the article Talk pages it belongs on. These three names have since become articles, and no note followed them.
106	2026-08-11	structural	low	Talk:Alompe, Talk:English language	Two Talk-page threads are still open on questions the operator has already closed, one of them for twenty-five days. Nothing revisits a Talk page when a verdict settles what it records, so all twelve live Talk pages — the 19 in the namespace, less 2 redirects and the 5 queued for deletion under deferred #58 — were read against the flagged-items table and the Dismissed flags list. Ten carry nothing but live questions. Two have drifted. Talk:Alompe says the native spelling Αλουμπε "follows Hertia's current orthography, which is itself under review (routine report, voice flag)" — written 2026-07-17 at 08:18. The review it names is #12, and it closed two days later: on 2026-07-19, in revision 1216 of this page, the operator rejected it — the Greek-letter native names on Hertia, Alompe and the Hertic pages are "fully intentional in-world orthographic style". There is no review, and there has not been one since. Talk:English language opens "The infobox's family field is left empty (isolate) — decide whether the Cargo Languages row should store a queryable value (e.g. "Isolate") instead". There is no Languages row to decide about: the encyclopedia's Cargo was retired on 2026-07-30 (dismissed #43 and #59), Template:Infobox language carries neither a cargo_declare nor a cargo_store, and a query against Languages throws the same MWException as Continents, LanguageFamilies, Polities and FormerCountries — re-verified this run, with Lexemes still the one table that answers.	Two edits, and both close a thread rather than open one. On Talk:Alompe, strike the second bullet, or replace it with the ruling it is waiting for — that the Greek-letter native spellings are settled in-world orthography per #12. Its first bullet stays: the map node recorded "until a settlement infobox exists to hold it" is still live, because no settlement infobox exists. On Talk:English language, strike the first bullet: an empty family field is correct for an isolate and there is no longer a table for it to feed. That page's other two bullets — the registry codes, and the pointer to Talk:Engil — both still hold. Neither edit touches an article.	pending	New 2026-08-11. Not a re-opening of #12, #43 or #59: all three are settled and stay settled, and that is exactly the finding — the settlement never reached the Talk pages that were waiting on it. The other ten check out, including the two that look stale and are not: the placeholder eras of dismissed #8 are still open threads, because the operator's ruling confirmed the names are not canon rather than supplying replacements, and Talk:Engil's two questions are undecided by his own word. Recorded beside it rather than filed: Talk:Engil instructs that {{Infobox polity}} be added once Engil's details are settled, and #74 proposes renaming that template — but #74's fix leaves a redirect behind, so the instruction keeps working whichever way that row goes.
107	2026-08-12	structural	medium	306 Dictionary: entries (Besoh)	The two languages that carry both headings order them opposite ways, and the one that keeps to the wiki's own skeleton is the smaller. Yezur Wiki:Dictionary structure fixes the order inside a language section and states it as a rule — "only the headings that apply, but always in this order and nesting" — with Alternative forms, then Etymology, then Pronunciation, then the part-of-speech heading. Ancient Hertic keeps it: all 212 of its sections put Etymology above Pronunciation, as does the wiki's one Pjany entry. Besoh inverts it, without exception: of its 1,103 sections, 306 carry both headings and all 306 put Pronunciation first — not one Besoh section is written in the documented order. The other 790 Besoh sections carry a pronunciation and no etymology, so the question does not arise in them. Nothing on the wiki records the inversion: Dictionary:About Besoh promises that "each entry's Pronunciation section gives the estimated spoken form" without saying where the section sits, Yezur Wiki:Conventions is silent on entry layout, and the structure page names no exception. The split is by language, not by author or by date — every Besoh section one way, every Hertic section the other, with no mixed page anywhere.	One of the two bodies has to move, and which one is the operator's call. (a) Bring the entries into line — in the 291 sections that carry a single Etymology, move the Pronunciation block below it. The move is as mechanical as an edit gets: all 1,096 Besoh Pronunciation sections are exactly one line, * {{auphen|word|YCQ-BE}}, with nothing else beneath the heading. (b) Bring the skeleton into line — record on Yezur Wiki:Dictionary structure that Pronunciation precedes Etymology, and the 212 Ancient Hertic sections become the deviation instead. (c) Record the order as a per-language choice and neither body moves. The remaining 15 sections are the same case under any ruling: they head their senses Etymology 1 and Etymology 2, where a pronunciation shared by both stands above them for the reason Wiktionary's own layout gives, and they are best left as they are.	pending	New 2026-08-12, from the first heading-structure pass the routine has run over the Dictionary — 2,415 entries, every language section's headings read in sequence and checked against the skeleton. Recorded beside it: two headings in daily use appear nowhere in that skeleton at all — Declension (245 sections) and Conjugation (141) — but their placement is consistent, directly under the part-of-speech heading, so no row is filed for them; the edit that settles this one could name them in passing. Update 2026-08-18: the balance has tipped decisively, and by a language that had no entries yesterday. All 500 Gaillean entries put Pronunciation above Etymology, without a single exception — 490 with one etymology, 10 with numbered ones. The wiki's split is now 806 sections in Besoh's order against 212 in the documented one: the form Yezur Wiki:Dictionary structure actually prescribes is kept by Ancient Hertic alone and by the single Pjany entry. That does not decide the row, but it changes what the three answers cost — (a) now moves 806 sections rather than 306, (b) now records what five sixths of the wiki already does, and (c) would have to explain why two Herta-Olaic languages order it opposite ways. A verdict is cheaper now than it will be after the next language. Update 2026-08-19: both totals moved by one and the shape did not — 807 sections in Besoh's order against 213 in the documented one (212 Ancient Hertic sections and the single Pjany entry). The one new section is Dictionary:soǵob, written this morning under #89, and it was written Pronunciation first like every other Besoh entry: 1,104 Besoh sections now, 307 of them carrying both headings, all 307 inverted, still without an exception anywhere. Re-counted 2026-08-20: the split is no longer 306 to 213. The 500 Gaillean sections carry Pronunciation above Etymology as well, so the deviant body now stands at 807 sections (Besoh 307, Gaillean 500) against 213 that keep to the skeleton (Ancient Hertic 212, Pjany 1). Nothing changed in this window to cause it — the Gaillean entries were written that way in August and the language was created after this row — but (a) is now four times the work it was when it was proposed, and the majority has moved to the other side.
108	2026-08-12	structural	low	42 Dictionary: entries (Proto-Andusian)	Forty-two entries carry a heading the operator's own ruling retired — and they carry it for a relation the ruling does not describe. Yezur Wiki:Dictionary structure records the ruling: one ====Related terms==== heading with labelled bullets "in place of Wiktionary's heading per relation", and it names ====Derived terms==== among the four headings it replaces. The ruling is otherwise kept perfectly: 81 Related terms sections hold 89 bullets, every one of them labelled — Compare: 38, Antonym: 32, Coordinate term: 14, Near-synonym: 4, Derived term: 1 — and there is not one unlabelled bullet in the Dictionary. But 42 Proto-Andusian sections head a list ====Derived terms====, holding 75 bullets, and the two lists are not the same thing: every one of the 75 is a {{desc}} line pointing into a daughter language — Proto-North-Andusian, Proto-South-Andusian, Proto-Unterian — while the one ruled Derived term: bullet, on Dictionary:ka, points at a Proto-Andusian word. The heading is used consistently: all 42 place it immediately before Descendants (41 of them; Dictionary:fiq, which has no Descendants section, ends with it), no bullet repeats a line from the Descendants list beneath it, and all 75 targets exist. Nine of the 42 carry both headings.	Two clean answers, and the routine will not choose between them — the choice states what the Dictionary means by a derived term. (a) Fold and drop — rewrite each bullet as * Derived term: {{desc|…}} beneath the section's Related terms heading, creating that heading where the section has none, and delete the Derived terms heading: 42 edits, after which the ruling holds without exception. (b) Document the heading — add Derived terms to the skeleton on Yezur Wiki:Dictionary structure as the cross-language list, between Related terms and Descendants, which is exactly where all 42 already sit: one page edit, no entry touched.	pending	New 2026-08-12, from the same heading pass as #107. Distinct from #87: that row is about pointers that resolve to nothing, and all 75 of these resolve. A 43rd page carries the heading and is left out of every count above — Dictionary:ɓolo, written unlinked because it is one of the 37 implosive-title pages in the deletion queue under deferred #70: it holds a delete tag, a stray Derived terms heading and one bullet, with no language section at all.
109	2026-08-12	structural	low	Dictionary:fiq	One line is written twice. The Derived terms list of Dictionary:fiq — the Proto-Andusian bound root covering — gives * {{desc|YCP-XC|fofiʔi}} on two consecutive lines, so a two-item list names the same Proto-Unterian word twice and the rendered entry prints Proto-Unterian: fofiʔi one line under the other. It is the only one on the wiki: 2,784 bullets were read across the Dictionary, grouped by the heading each sits under, and this is the single repeated line among them.	Delete the second of the two lines. Nothing else on the page changes.	pending	New 2026-08-12. Independent of #108: the duplicate goes whichever way that row is settled — folded into one labelled bullet under (a), deleted where it stands under (b) — so this row can be approved on its own. Dictionary:fiq is also the one page of the 42 that has no Descendants section, its list of derivatives being all the onward material it carries.
110	2026-08-13	structural	medium	206 Dictionary: entries (Proto-Herta-Olaic), Dictionary:gurə	One job, three notations — and the split is by language family, not by page. Every Descendants list on the wiki was read: 1,271 bullets holding 1,273 links. The Andusian layer writes all 1,064 of its bullets as * {{desc|code|word}}. The Proto-Herta-Olaic layer writes all 206 of its bullets by hand — * Ancient Hertic: {{l|YAQ-HE|word}} — spelling out the language name that Template:Desc reads from Module:Languages. The single Proto-Njeshan bullet, on Dictionary:gurə, uses a third form and a third template: * Pjany: {{d|geúrl|YBS-PJ}}, word first. No page mixes two of them, and the rendered output of all three is identical — test-parsed side by side, the hand-written Hertic bullet and {{desc|YAQ-HE|αμπαν}} emit byte-for-byte the same HTML, section anchor and all. What differs is maintenance, and Template:Desc's own documentation names it: the display name comes from the registry, "so renaming a language is a single registry edit, never a page sweep". Renaming Ancient Hertic today means a 206-page sweep.	207 edits, and not one changes what a reader sees. Replace each hand-written bullet with * {{desc|YAQ-HE|word}}, and Dictionary:gurə's with * {{desc|YBS-PJ|geúrl}}. One page needs a second decision: Dictionary:xmɑtʲe: (Proto-Herta-Olaic, the genderless third person) lists three Ancient Hertic reflexes on one line — * Ancient Hertic: {{l|YAQ-HE|μα}}, {{l|YAQ-HE|εμι}}, {{l|YAQ-HE|του}} — the only Descendants bullet on the wiki naming more than one word; under {{desc}} it becomes three bullets, which is how every other multiple-reflex entry is already written.	pending	New 2026-08-13, from the first reciprocity pass over the Dictionary — see the run notes. Not a rendering fault: nothing is broken today, and the row would be low but for the sweep it stores up. Recorded beside it: accepting the notation makes the Proto-Herta-Olaic layer's bookkeeping perfect — 208 descent claims, 208 descendant links, zero unanswered in either direction, the only layer on the wiki with a clean sheet. If the operator would rather keep the hand-written form, the row closes as rejected and the practice should be written into Yezur Wiki:Dictionary structure, which currently documents neither form. Update 2026-08-18: the two notations now sit on the same page, in the same list, which they did not when this row was filed. The 210 Proto-Herta-Olaic entries that gained a Gaillean reflex this window all write it as * {{desc|YAQ-GA|word}} — the form this row proposes — and 205 of those pages already carried a hand-written Ancient Hertic bullet, so a reader now sees "* Ancient Hertic: {{l|YAQ-HE|…}}" directly above "* Gaillean: …" produced by the template. The newest work has answered the row in practice; carrying it out would now make 206 hand-written bullets match the 226 template ones beside them, rather than converting an isolated layer. The count is otherwise unchanged: 206 hand-written bullets and Dictionary:gurə's third form.
111	2026-08-13	structural	medium	Dictionary:ɔŋ	A fifth Descendants list left behind by the rope relocation, on the page nobody checked. Dictionary:ɔŋ (Proto-South-Andusian, rope, cord; vine) gives its Besoh descendant as {{desc|YCQ-BE|on}}. Dictionary:on holds no rope sense: it is the adposition at, by, on, inherited from Dictionary:oŋ — a different Proto-South-Andusian etymon, whose own Descendants list names on correctly. The Besoh reflex of rope is one page away and is written: Dictionary:one, glossed rope, cord; vine, liana, whose etymology reads "From on (“rope”, now archaic), extended to escape its homophone, the locative particle." The link is blue, so no maintenance page can see it; it simply lands on the wrong word.	One line. On Dictionary:ɔŋ, change the descendant to * {{desc|YCQ-BE|one}}. Nothing is asserted that Dictionary:one does not already say about itself.	pending	New 2026-08-13. Not a re-opening of #84: that row's Notes record checking Dictionary:oŋ "for the same residue after the rope homograph moved" and finding it correct — and it is correct, because it is the locative etymon. Dictionary:ɔŋ, with the open vowel, is the rope etymon, a separate page, and is the one holding the residue. A second question the operator may want to settle in the same edit, filed here rather than as its own row: Dictionary:one names its source as plain italic on with no {{m}} and no Proto-South-Andusian step, so the chain Dictionary:ɔᵑg → Dictionary:ɔŋ → Dictionary:one is unlinked at its last step from both ends; and the archaic Besoh on the etymology cites is not a sense Dictionary:on carries.
112	2026-08-13	factual	medium	Dictionary:bë, Dictionary:be, Dictionary:gön, Dictionary:gö	Two words say they exist to escape a homophone that is still sitting where they left it. Four entries explain themselves as a form split off to avoid a clash, and two of the four are consistent: Dictionary:mium (hail) fronted "to escape mum (“scratch”)", and Dictionary:mum's Besoh section carries scratch and nothing else; Dictionary:one (rope) extended "to escape its homophone, the locative particle", and Dictionary:on carries the locative and nothing else. The other two are not. Dictionary:bë is glossed right and opens "From Proto-South-Andusian be (“right”), the vowel lengthened to escape its homophones" — while Dictionary:be's Besoh section still carries Etymology 3, an adjective glossed right(-hand); the strong side. Dictionary:gön is glossed spear and opens "From gö (“spear”), extended with -n to escape its homophones" — while Dictionary:gö still carries Etymology 3, a noun glossed spear, lance. Lexemes therefore holds two Besoh adjectives for right and two Besoh nouns for spear. And the ancestors follow the source, not the escape-form: Dictionary:be's Proto-South-Andusian right section lists Besoh be, and Dictionary:goː's spear section lists Besoh gö.	The reading is the operator's; both readings are then mechanical. (a) The escape completed — delete Etymology 3 from Dictionary:be's Besoh section and from Dictionary:gö, and repoint the parents to {{desc|YCQ-BE|bë}} and {{desc|YCQ-BE|gön}}. This is the pattern the wiki already used for kunke → Dictionary:kiunke, where the superseded page was blanked and queued. (b) Both forms stand — keep the older sense, mark it archaic or dialectal on its own entry so the escape sentence stops contradicting it, and list both descendants on each parent.	pending	New 2026-08-13, from the same reciprocity pass as #110. Not a re-opening of #84 or #85: those are links pointing at the wrong page or a sense the parent does not have, and both are answered by editing the parent; this is two live entries on the same language contradicting each other, and no edit to a parent settles it. Invisible to the duplicate test that runs every night: it matches on (word, language, part of speech) and these are two spellings of one sense, which no uniqueness check on the table can see. Not a criticism of either coinage — the escape pattern is well attested here, and the two entries that get it right show exactly what completing it looks like.
113	2026-08-14	factual	medium	10 Dictionary: form entries (Besoh), against Dictionary:mum, Dictionary:mium, Dictionary:on and Dictionary:one	Two of the four homophone escapes of #112 are complete on the lemma page and unfinished one level down: the old word's inflected forms are still on the wiki, still declining a lemma that no longer holds that sense. Dictionary:mium (hail) says it was fronted "to escape {{m|YCQ-BE|mum}} (‘scratch’)", and Dictionary:mum's Besoh section duly carries the verb scratch and nothing else — no noun, and a Conjugation table rather than a Declension one. Beneath it stand five noun-form entries — Dictionary:mumbi, mumon, mumu, mumubi and mumu'on — each reading "…of mum (‘hail’)", a Besoh noun the Dictionary elsewhere denies exists; the five forms of mium are written in full beside them, so the paradigm stands on the wiki twice. Dictionary:one (rope) says the same of itself — extended from on "to escape its homophone, the locative particle" — and five more form entries name on and gloss it "rope": Dictionary:onbi, onon, onu, onubi and onu'on. Dictionary:on is the adposition at, by, on; it holds no rope sense and no table, so a reader following any of the five arrives at the wrong word. The two halves fail differently, because the escape changed only the singular stem. The plural of one is still onu, so onu, onubi and onu'on are the right strings under the wrong name — they are Dictionary:one's own live table cells — while onbi and onon are the archaic singular and are generated by nothing at all, the dative and locative singular of one being onebi and one'on, both written and both correct. No existing check can see any of it: every link involved is blue, the Lexemes duplicate test matches on (word, language, part of speech) and so cannot see two spellings of one sense, and all ten pages pass the 2026-08-12 anchor pass and the 2026-08-13 reciprocity pass alike — neither of which read a non-lemma entry.	Mechanical in both halves, and the routine invents nothing — every judgement below is read off a page that already exists. (a) The three misnamed cells: change {{l|YCQ-BE|on}} to {{l|YCQ-BE|one}} on Dictionary:onu, onubi and onu'on, one word each. The gloss "rope" then matches the lemma it points at, and the three become what Dictionary:one's table already says they are. (b) The five stranded mum forms and the two stranded on singulars: tag them {{delete}} with the reason stated, exactly as the fifty superseded plurals of #81 and the two kiunke respellings already on the queue are tagged. The routine holds no delete right, so tagging is the whole of what it can do, and an administrator clears them with the rest. One of the seven is edited rather than tagged: Dictionary:mumu also carries a Proto-Unterian noun (sea-foam), so only its Besoh section goes and the page stays. The alternative is a claim about Besoh and the routine will not make it — that mum does still hold the noun hail and on the noun rope, in which case the ten entries stand, both lemma pages need the sense restored with a table under it, and Dictionary:mium and Dictionary:one need their escape sentences rewritten.	pending	New 2026-08-14, from the first inflection-reciprocity pass. The ten titles are written unlinked: six of them — mumbi, mumon, mumubi, mumu'on, onbi and onon — are on Special:LonelyPages today, and a link from this Report would take them off it and cost the count its meaning. That those six are lonely is this finding seen from the other side: of the 797 Besoh form entries on the wiki, exactly six have no incoming link, and they are exactly the six that no live table generates. Special:LonelyPages has held that evidence since the routine first read it, and no run had asked it this question. Relation to the rows already on the table: #112 names all four escapes and rules on the two unfinished ones (Dictionary:bë and Dictionary:gön); this row is about the two it called consistent — which they are, on the lemma pages. #111 is the same rope residue in a third place, a Descendants bullet on Dictionary:ɔŋ, and #84 is a fourth. Approving #111 does not touch these five pages, and approving this row does not touch that bullet. Dictionary:bë leaves no residue of this kind, Besoh be being an adjective and taking no table.
114	2026-08-14	structural	medium	Dictionary:hobi, Dictionary:ho'on, Dictionary:hu'on, Dictionary:hummubi, Dictionary:hummu'on	Five non-lemma entries are linked by a live inflection table that they do not acknowledge. Where two Besoh senses inflect into one string the form entry is meant to list both, and across the Dictionary it does: all 59 pages on which a vowel-final noun's locative singular and locative plural collapse into one form state both slots, without exception, and Dictionary:simubi and Dictionary:simu'on carry a line each for sim (wood) and simu (timber), as Dictionary:śimubi and Dictionary:śimu'on do for śim and śimu. Six cells break that rule. Dictionary:ho holds two Besoh nouns — egg under Etymology 1, stone under Etymology 2 — each with its own Declension table. Dictionary:hobi is the dative singular of both and names only egg. Dictionary:ho'on is the locative singular of both, and of Dictionary:ho' (guts) besides, which it does name — and it again omits stone. Dictionary:hu'on is the locative plural of both ho senses and names neither, speaking only for Dictionary:hu' (dog). Dictionary:humme (cloth) is the third instance of the sim/simu shape and the only one left incomplete: its plural is hummu, so its dative and locative plurals are Dictionary:hummubi and Dictionary:hummu'on, and both speak only for Dictionary:hummu (loom). Dictionary:hummu itself was written correctly — its Noun 2 is the plural of humme — so the pattern was begun on that pair and stopped two pages short. What a reader meets: following the dative plural in Dictionary:humme's table lands on a page that says "Dative singular of hummu (‘loom’)", a different word with a different meaning, and following either ho table's locative plural lands on a page about a dog.	Six lines added to five pages, in the wording those pages already use, and nothing removed. On Dictionary:hobi: # Dative singular of {{l|YCQ-BE|ho}} (“stone”). On Dictionary:ho'on: the same for the locative singular of ho (stone), beside the three lines already there. On Dictionary:hu'on: two lines, the locative plural of ho (egg) and of ho (stone). On Dictionary:hummubi: # Dative plural of {{l|YCQ-BE|humme}} (“cloth”). On Dictionary:hummu'on: # Locative plural of {{l|YCQ-BE|humme}} (“cloth”). Nothing new is asserted by any of the six: each is a cell Module:Besoh generates today, from a table already on the page, and each is copied from the shape Dictionary:simubi and Dictionary:simu'on already use for the identical situation. Order within each list follows those two — the shorter lemma first.	pending	New 2026-08-14, from the first inflection-reciprocity pass. The count is a floor, not a guess: the routine rebuilt Module:Besoh's morphology locally and then checked it against the live parser on every one of the 242 lemma headwords that carries a table — 242 of 242 identical, link for link — before comparing anything, so the six cells are proved rather than inferred. Four more cells of the same shape are deliberately not counted here, because #112 decides them: Dictionary:gö's Etymology 3 (spear) has a live Declension table whose dative singular, dative plural and locative singular and plural land on Dictionary:göbi, gö'ubi and gö'on, all three of which speak only for gö (louse). Settle #112 the first way and that table goes, taking the four cells with it; settle it the second way and they want the same four lines this row proposes, which the routine can carry out in the same pass. Beyond these and the ten pages of #113, no entry on the wiki states a slot its table does not generate, and no cell lands anywhere unclaimed — the two rows are the whole of the disagreement.
115	2026-08-15	structural	low	7 Dictionary: entries (Besoh)	Seven Besoh entries are the only ones the Dictionary never says how to pronounce — and the module pronounces all seven. Dictionary:About Besoh opens its account of the sound system with a promise: "The spelling is conservative: Besoh is not pronounced as written. Each entry's Pronunciation section gives the estimated spoken form." Of the language's 1,103 sections, 1,096 keep it. The seven that do not are all bound forms — -ke, -me, -so, -va, bi-, ro-, so'- — which reads as a class exemption until the eighth affix is read. Dictionary:bon- is a Besoh prefix and it does carry one, * {{auphen|bon-|YCQ-BE}}, in the ordinary Besoh position above the etymology. So the Dictionary holds eight Besoh affix entries and pronounces one. The module is not the obstacle, and it was built for exactly this case: test-parsed live, all seven resolve and none files into Category:Auphen errors — and a trailing hyphen visibly suppresses the word-final laws, so bon gives /bõ/ while bon- gives /boŋ-/, which is the law Dictionary:About Besoh states in its own words: "Final n survives only as nasality on the vowel … the [ŋ] returns under suffixation". Module:Auphen already knows what a Besoh affix sounds like; no entry asks it.	Seven one-line insertions, and the line is already written for each. Add to each of the seven a ===Pronunciation=== heading with the single bullet * {{auphen|WORD|YCQ-BE}}, WORD being the entry's own headword, placed above the ===Etymology=== heading, where Dictionary:bon- and every other Besoh entry put it. The routine asserts nothing: the rendered form is whatever the module returns, exactly as on the other 1,096.	pending	New 2026-08-15, from the first pass the routine has made over the pronunciation layer — all 1,309 {{auphen}} calls in the Dictionary read and checked against the section holding them. Stated against the row: of the seven, only so'- would render differently from its spelling (/soʔ-/); the other six come back as written, so six of the seven edits buy consistency rather than information, and that is a fair reason to close this rejected and record the affixes as a deliberate exception on Yezur Wiki:Dictionary structure instead. It does not presume #107: if that row is settled by moving Besoh's Pronunciation below Etymology, these seven should be written in whichever order it fixes, and the same seven edits serve. Not a dormant corner — four of the eight affixes are named from other entries' etymologies (-ke twice, -so three times, -me and bon- once each).
116	2026-08-15	structural	low	105 Dictionary: entries (71 Besoh, 32 Proto-Andusian, 2 Ancient Hertic)	The Dictionary records a word's descent from a foreign parent at both ends, and its formation from a parent in its own language at one end only. Every etymology naming a word in its own language section was read against what that word's page lists back: 144 such formations — Besoh 99, Proto-Andusian 43, Ancient Hertic 2. Three are answered, all three Proto-Andusian and all three by a Related terms bullet: Dictionary:ka names kaka as a Derived term:, Dictionary:bʰek names bʰebʰe and Dictionary:mɔ names moᵑg, both as Compare:. The other 141 are silent in that direction. 108 distinct entries are named as the source of a word formed from them — every one of the 108 is itself an entry on this wiki, with the sense the derivative claims — and 105 of them say nothing back. Besoh is the pattern entire: 99 formations — korso "the runner" from koro "run", vezen "of-cloud" from vez "cloud", bonkoro "far-running" from both — and across all 1,103 Besoh sections there is no Related terms section and no Derived terms section of any kind. The cause is structural, not careless, and #108 is where it surfaces: all 75 Derived terms bullets on the wiki are {{desc}} lines pointing into a daughter language, so the Dictionary has no established home for a formation inside one language — which is why the single page that tried to record one reached for the Derived term: label, used exactly once in the Dictionary.	flag-only, and it waits on #108. That row decides what ====Derived terms==== means; until it does, there is nowhere agreed to put these 141 bullets, and the routine will not invent a section. Once it is settled the addition is mechanical and purely additive — one bullet on each of the 105 silent bases naming the words formed from it, every one read off those words' own etymologies, no prose touched and nothing asserted that the derivative does not already state.	pending	New 2026-08-15, from reading the Dictionary's cross-reference layer in both directions. Distinct from #108: that row is about which heading carries which relation and would be satisfied by renaming headings without one new bullet; this is about the bullets that are absent under any heading. Recorded as not findings, from the same pass: the symmetric relations are perfect — of the 89 Related terms bullets, Antonym 32 of 32, Coordinate term 14 of 14 and Near-synonym 4 of 4 are reciprocated without exception, and only Compare: (38 bullets, 16 mutual) runs one way, which is what a compare-note is; and all 75 cross-language Derived terms bullets are answered by their target's own etymology, 75 of 75.
117	2026-08-17	markup	low	Template:D	The one English word the Dictionary does not have is the one its link template demonstrates on. Template:D §Usage shows three worked calls, each written twice — the escaped syntax, an arrow, then the live call: {{d|dog|YXS-EN}}, {{d|dog}} and {{d|dog|YXS-EN|dogs}}. The live half executes, so the documentation renders three links to Dictionary:dog, and no such entry exists. The Dictionary holds exactly one English entry, Dictionary:warm — already this template's most-used target, since the five entries whose definition line is a link to it (ak͡xo:wn, geúrl, gurə, ɲehits, εκωμε) each write that line as {{d|warm|YXS-EN}}. The prose asserts the entry too — "the English section of Dictionary:dog" — and Yezur Wiki:Dictionary structure repeats the same example and the same sentence, escaped there, so it creates no link of its own. The consequence is measurable, and it is singular. Every red link on the wiki was traced to its source this run: of the 60 titles on Special:WantedPages, Dictionary:dog is the only one that no live row, no dismissal and no item on the operator's to-do list accounts for, and the only one anywhere emitted by a template's own documentation. The other 59 are the four Ancient Hertic paradigms (#21), the five Tacchoan names (#57), Map:Conventions (#65), the twelve unwritten roster languages (#75), the two era pages (#8), {{tl}} (#79), the eleven affix titles (#87), Dictionary:soǵob (#89), the two map values (#33), and the ten lexical categories and ten module /doc pages on User talk:Yezur.	Three words, and the replacement is the template's own most-used target. Change the example word from dog to warm in all three bullets — {{d|warm|YXS-EN}}, {{d|warm}}, {{d|warm|YXS-EN|warmer}} — and the sentence to "the English section of Dictionary:warm". Both halves of each bullet move together, so syntax and rendering stay in step; the documentation then demonstrates a live link into a real section rather than a red one, and Dictionary:dog leaves Special:WantedPages. The third call needs display text: warmer is proposed, is display text only, and asserts nothing about the entry — any string serves. (b) The alternative is to escape the live halves, as #102 asks for Template:Desc, but that costs the demonstration Template:L and Template:M both keep, so (a) is the better answer here. The escaped copy on Yezur Wiki:Dictionary structure can follow in the same pass; it changes nothing a reader can click.	pending	New 2026-08-17, from the first end-to-end read of the wiki's link graph. Not a re-opening of #102: that row is about escaping, and it names Template:D among the templates that escape correctly — which is true, it shows syntax and live output side by side. This row is about where the live call lands, a question #102 asked only of its own template, where it checked and found Dictionary:weki present. Recorded and not flagged: every other worked example on the wiki lands on a page that exists, which the accounting above proves — a documentation example pointing at a missing page appears on Special:WantedPages with its template named as the source, and no other does. The title is written unlinked here so this Report does not become a second source of the red link the row is about.
118	2026-08-17	structural	medium	Dictionary:Main Page	The Dictionary's front page offers three ways into the words, and all three lead out of the Dictionary. Dictionary:Main Page is one click from every page on the wiki — MediaWiki:Sidebar carries it under Yezur — and Main Page links it in its closing panel. Its §Ways into the words holds three cards, and between them nine links: Ancient tongues to Ancient Hertic language and Gaillean language; Living tongues to Lonish language, Nichana language, Pjany language and Uu language; Families & roots to Category:Language families; and a closing line to Main Page and Map:Main Page. Not one of the nine points at a Dictionary page — not at an entry, not at one of the five Dictionary:About … appendix pages. Nothing comes back from the other side either. All 4,846 links on the wiki were read this run: exactly one runs from the encyclopedia into the Dictionary (Main Page to this page), and fourteen run the other way, all of them from six pages — this one and the five appendix pages. No language article links to the Dictionary section of its language; no entry links to the article of the language it is written in. The category route is closed as well: of the 105 categories that hold an entry, every one is red — that tree is the operator's to-do item #18 — so the only live categories any word sits in are Category:Pages flagged for deletion (89 entries) and Category:Ancient Hertic entries needing etymology (3). Measured end to end: following links and category membership through content pages from Main Page reaches 70 pages — all 32 articles, 37 categories, and Dictionary:Main Page — and not one word. Counting the deletion queue as a road reaches 859 more; counting the wiki's own machinery (this Report, a template's documentation) reaches 1,198 of the wiki's 2,622 pages.	Half of it can be done today; the other half waits on to-do #18. (a) Now: give the cards a target inside the Dictionary. The five appendix pages exist and are the Dictionary's own writing about its languages — Dictionary:About Proto-Andusian and its three daughters document the proto-languages, their sound laws and their derivation, which is what Families & roots promises, and Dictionary:About Besoh is an attested language, which is what Living tongues promises. Five links, asserting nothing that is not already on those pages. (b) Waits: the door a dictionary front page really wants is a lemma category per language, and none exists yet; when the tree of to-do #18 is built, each card can point at its Category:X lemmas and this row closes properly. (c) A third answer the routine will not choose on its own: leave the cards pointing at the language articles and have those articles link their Dictionary section instead, which closes the same gap from the far end and touches eleven pages rather than one. Naming any new category remains the operator's, per Yezur Wiki:Conventions.	pending	New 2026-08-17, from the first end-to-end read of the wiki's link graph — 4,846 links and 2,553 category memberships, in both directions. Distinct from #95, which is Main Page's Explore the world panel and its two cards that carry no link at all: here every card carries a link and every link reaches a real page — the fault is where they go. Distinct from #77, which is about the appendix pages having no category, and whose notes already record their being unlinked as not a finding. Does not re-open to-do #18: the category tree stays where the operator put it, and is named here only as the blocker on half the fix. Recorded beside it: Special:Random sits in the sidebar and does reach entries, and Special:AllPages and the search box list them — this row is about following links and categories, which is how an encyclopedia is read.
119	2026-08-18	voice	high	Dictionary:About Middle Airananwe	The wiki's strictest rule is broken by one clause on a page created this window. Dictionary:About Middle Airananwe §Gender closes its account of the collapse of the neuter with: "The neuter split by stem, as Latin's did: consonant-stem neuters joined the masculine, vowel-stem neuters went feminine." Yezur Wiki:Conventions §The Earth test forbids exactly this — a real Terran language named in article namespace, and named as the authority for how a Yezuri language changed. It is the only in-frame Earth reference on the wiki: all 746 pages touched this window were read for the forty-odd names that would breach the rule, and every other hit is out of frame — three in Lua comments on Module:Head ("Latin-style class category"), one on Module talk:Auphen, and the escaped comparison passage on Yezur Wiki:Dictionary structure, which is out of frame by the namespace table. The page does not need the comparison to work: the split it describes is stated in full by the clause after the colon.	Delete four words. "The neuter split by stem: consonant-stem neuters joined the masculine, vowel-stem neuters went feminine." Nothing else in the sentence changes, no claim is lost, and the mechanism the page is documenting is stated exactly as before. If the parallel is wanted as a note on how the change is to be read, [[Talk:Dictionary:About Middle Airananwe]] is where it belongs — that is the division Yezur Wiki:Conventions draws, and the one this Report's own out-of-frame commentary keeps.	pending	New 2026-08-18. Filed high not because it is large but because of which rule it is: it is the line the convention set calls the strictest application of the no-Earth rule, and it is four words of a page that is otherwise the best-documented sound-law account on the wiki. Not a criticism of the page: its nine developments, its orthography and its gender table were checked against the Gaillean entries and against Module:Auphen/YAQ-GA, and hold. Recorded as not a finding beside it: IPA appears on Dictionary:About Gaillean and throughout the Dictionary, and is settled practice no run has questioned. Resolved 2026-08-20, and not by a verdict: Dictionary:About Middle Airananwe was rewritten in full at rev 7475 for the ratified design, and the §Gender section is gone altogether — the language now has no gender to lose, so the sentence carrying the comparison went with the section. Checked against the live page and against a wiki-wide sweep for 35 real-world terms: no in-frame page uses any of them. The row needs only a verdict to clear.
120	2026-08-18	structural	medium	8 Dictionary: entries (Gaillean) — annawe, bell, beule, kelte, kintuz, mbezwettuz, mbila, vite	Eight of the new Gaillean pages say the same thing twice, and the Cargo table is what shows it. The Gaillean sections hold 512 {{head}} calls and Lexemes holds 504 rows for the language — and the eight missing rows are not lost, they are collapsed: on each of these eight pages the same entry is written out twice, under ===Etymology 1=== and ===Etymology 2===, identical in origin, part of speech, gloss, class and definition line. Dictionary:annawe gives fire twice from the same Proto-Herta-Olaic etymon; Dictionary:mbila child twice; Dictionary:vite water twice; Dictionary:kintuz worm, Dictionary:mbezwettuz wind, Dictionary:beule orphan and Dictionary:bell for the same way. Dictionary:kelte is a triple — male, then man and man again, the last two byte-identical. Yezur Wiki:Dictionary structure reserves the numbered-etymology pattern for "where one spelling covers several unrelated origins"; here the origins are not merely related, they are the same words. Cargo stores one row per distinct field-set, so the duplicate is invisible in the data and visible only on the page. The declension pass then wrote a table per section but placed them all in Etymology 1, so seven of the eight also carry two identical {{YAQ-GA-decl}} tables in a row — Dictionary:kelte carries three — and none in the section below.	Delete the duplicate, and the surplus table with it. On each of the six plain noun pages: remove the second Etymology block entirely, remove the second {{YAQ-GA-decl}} call, and demote the surviving headings one level (====Noun==== to ===Noun===, =====Declension===== to ====Declension====) since the page no longer has numbered etymologies — the shape every other single-etymology Gaillean entry already has. Dictionary:bell has no table: remove its Etymology 2 and demote. On Dictionary:kelte, which keeps two genuine senses, remove Etymology 3 and the third table only, leaving Etymology 1 (male) and Etymology 2 (man) numbered as they are. Nothing a reader can learn is lost — every removed block is a copy of one that stays — and after the pass the head count and the Lexemes row count agree at 504.	pending	New 2026-08-18, from reconciling {{head}} calls against Lexemes page by page: these eight are exactly the shortfall, and no other page on the wiki shows one. Distinct from #121, which is the two pages where the doubled table sits over senses that are not duplicates. Recorded as not a finding: the category machinery counts these correctly already — every one of the sixteen Gaillean lemma and word-class categories reconciles page-for-page with the entries (lemmas 500, nouns 311, verbs 82, adjectives 44; F 93, M1 71, M2 31, N1 44, N2 24, MX 20, FX 15, NX 6) — because a category counts pages and cannot double-count a page against itself. Note for #124: the entry total quoted on both Dictionary:About … pages is 512, counted the same way as the head calls, so carrying out this row moves it to 504 and those figures should move with it. Update 2026-08-19 — carried out, and the row now reads clean. Thiorosan merged all eight pages this morning at the operator's request (revs 7428–7437), each now a single un-numbered Gaillean section with one {{head}}, one sense list and one declension table; kelte keeps male and man as two senses of one entry rather than two numbered etymologies. Re-checked page by page against the table: zero duplicated head calls remain anywhere on the wiki, and the head count and the Lexemes row count now agree exactly, per language and in total, at 3,239. One correction to this row's own projection: the agreed Gaillean figure is 501, not 504. Eleven head calls went rather than eight — kelte carried three and lost two — and #121's ieca and peramme, merged in the same pass, collapsed two further distinct Lexemes rows, which this row's arithmetic could not have anticipated because those two were not duplicates. The note this row leaves for #124 stands with 501 in place of 504. The row itself is left in the table, pending, for the verdict that clears it.
121	2026-08-18	structural	medium	Dictionary:ieca, Dictionary:peramme	Two more Gaillean pages carry a doubled declension table, and on these two the senses are real. Unlike the eight of #120, Dictionary:ieca and Dictionary:peramme state genuinely different senses in their numbered etymologies — ieca: female (noun), female (adjective), then woman from a named Proto-Herta-Olaic etymon; peramme: goat, then he-goat. But the declension pass placed both of each page's tables in Etymology 1, so ieca shows two identical {{YAQ-GA-decl|class=F}} calls under female and none under woman, and peramme two class=M1 calls under goat and none under he-goat. Both of the later sections are nouns, both carry a class in the headword line, and both therefore want the table sitting one section above them. The second half is a question, not a defect. The structure page reserves numbered etymologies for "several unrelated origins", and it is not obvious that female/woman or goat/he-goat are unrelated: iecas first two sections carry {{etym-missing}} and only the third names an etymon, and both of perammes carry it. Read as one origin, each pair is one entry with two numbered senses.	'(a) Mechanical, and independent of the question: move the second table down — on Dictionary:ieca, delete the duplicate class=F call under Etymology 1 and add a Declension heading with {{YAQ-GA-decl|class=F}} under Etymology 3; on Dictionary:peramme, the same with class=M1 under Etymology 2. Two edits, no rendered table lost, and each sense then shows its own paradigm. (b) Flag-only, and the operator's: say whether these are separate origins. If they are not, the merge is also mechanical — one part-of-speech section per pair with two # lines (# goat / # he-goat), one table, and iecas adjective kept apart from its noun as it already is — and (a) becomes unnecessary. (a) is safe either way and can go first.	pending	New 2026-08-18. Split from #120 deliberately: that row deletes text, this one moves it, and running one verdict across both would risk deleting a live sense. Recorded as not a finding: the class named in each doubled call matches the class in the headword line above it — here, and on all 283 pages carrying a table, with zero mismatches wiki-wide. Update 2026-08-19 — carried out, and as (b) rather than (a). Dictionary:ieca (rev 7431) and Dictionary:peramme (rev 7436) were merged this morning rather than having their second table moved down: each is now one un-numbered section — ieca with a Noun carrying both female and woman and its Adjective kept apart as it already was, peramme with one Noun carrying goat and he-goat — and each shows exactly one declension table, of the class its headword line names. The operator's reading is therefore that neither pair is a pair of separate origins, which is the half of this row the routine would not answer. Both pages parse clean and neither lost a sense. Left pending for the verdict that clears it.
122	2026-08-18	structural	low	Category:Gaillean entries needing etymology (276 members, title unlinked — see Notes), Template:Etym-missing	The review queue the template promises does not exist for the language that fills it. Template:Etym-missing files each page it is called on into Category:<Language> entries needing etymology, and its own documentation calls that "a reference queue to look over later (not flagged by the daily routine)". For Ancient Hertic the category is live: Category:Ancient Hertic entries needing etymology exists, holds 3 pages, carries a one-line description and sits in Category:Dictionary maintenance. For Gaillean it is red. 276 of the 500 new entries call the template, so Category:Gaillean entries needing etymology is now the largest wanted category on the wiki that is not a lemma or word-class category — 6th of 125 on Special:WantedCategories, above every category for a real part of speech. The 276 pages are filed correctly, and filed into nothing.	One page, copying a page that already exists. Create Category:Gaillean entries needing etymology with the Ancient Hertic wording and one word changed: "Gaillean Dictionary entries whose etymology has not yet been established, filed here by {{etym-missing}} for review. A reference queue — not flagged by the daily maintenance routine.", in [[Category:Dictionary maintenance]]. The routine invents no name and no placement: both are read off the sibling. Category:English entries needing etymology (1 member, Dictionary:warm) is the same case and the same edit if it is wanted.	pending	New 2026-08-18. Not an instance of the wanted-category class that to-do #18 covers: those are the lemma and word-class categories {{head}} emits, which the operator has taken as one decision. This is a maintenance category whose sibling already exists as a decided precedent, so the convention set's reservation of new category names is satisfied by copying the name that was already chosen. Not a request to act on the 276 entries: the queue exists to be looked at later, and this row asks only that it be somewhere.
123	2026-08-18	factual	medium	Dictionary:About Middle Airananwe, Herta-Olaic, List of languages	A page created this window says the language has no native name; the two lists that name it disagree. Dictionary:About Middle Airananwe states it flatly, on a line of its own: "Middle Airananwe has no recorded native name." Both hand-maintained rosters give one. The Members table on Herta-Olaic reads Middle Airananwe · Airannawê · Medieval; the Historical languages table on List of languages reads Middle Airananwe · Airannawê · Herta-Olaic · Medieval · YAR-AI · Ancestor of Airananwe. The two rosters agree with each other exactly and both predate this window. The form is also plausible against the new page's own sound laws — modern Airánanue beside the circumflex vowels the page introduces — which is what makes this worth a verdict rather than a guess: the name may be right and the sentence stale, or the name may be a placeholder that the sentence is retiring.	flag-only for the reading; both readings are then one edit. (a) If Airannawê stands: delete the sentence from Dictionary:About Middle Airananwe, or replace it with the name, as the page's opening paragraph already does for Gaillean ("Gaillean (Ghaellea, code YAQ-GA)"). (b) If the language has no native name: blank the Native name cell in both tables — an em dash, as List of languages already writes for High Nichana, Otjesi and Pises. The routine will not choose: a language's own name for itself is canon.	pending	New 2026-08-18, from the hand-maintained-lists check. The check otherwise passed: every other cell of the two rosters agrees with the article it summarises, Herta-Olaic's Members table holds all six of the family's languages with Gaillean correctly added, the family field of all 11 language infoboxes matches its family article (Herta-Olaic 2, Njeshan 7, Andusian 1, English none), and the five codes touched this window — YAQ-GA, YAR-AI, YAS-AI, YAR-GW, YAP-XX — are identical in Module:Languages, on List of languages and in the infoboxes. Recorded as not a finding: Gaillean language does not mention that Middle Airananwe descends from it, though both Dictionary:About … pages say so — an incompleteness on the article rather than a contradiction, and worth a sentence whenever that article is next touched. Re-checked 2026-08-20: the page was rewritten at rev 7475 and the contradiction survived the rewrite, in new words — the sentence now reads "No native name is recorded for it." in the opening paragraph rather than on a line of its own. Both rosters are unchanged and still give Airannawê. The row stands as filed; only the quotation needs reading against the new wording.
124	2026-08-18	factual	low	Dictionary:About Gaillean	The appendix page's own noun table is one noun out, and its last sentence about the irregulars stopped being true four hours after it was written. (a) §Nouns tabulates the six declensions with a count each — F 95, M1 77, M2 31, N1 46, N2 25, irregular 41. Five of the six are exact against the entries, counted the way the page counts (one per {{head}} call): F 95, M1 77, M2 31, N1 46, and MX+FX+NX = 20+15+6 = 41. N2 is 24, not 25 — the wiki is one -u noun short of what the page says it holds, and it is the only figure on either appendix page that does not reconcile. (b) The section closes "Their tables are not yet built" of the irregular classes. That was written at 12:39; at 14:25 Module:Gaillean gained the twenty masculine irregular stems and their paradigm, and all 20 MX entries have carried a rendering declension table since. The sentence is now true of FX and NX only.	(a) Either correct the cell to 24, or add the missing -u noun — the routine cannot tell which, because the page is the only record of the intended figure and the source lexicon is the operator's. If the count is simply wrong it is one character. (b) Rewrite the closing sentence to say that the masculine irregulars decline from stems recorded in Module:Gaillean and that the feminine and neuter tables are not yet built — which is what the module and the entries now show, and which is also the sentence #125 would retire. Both halves are one edit to one section.	pending	New 2026-08-18. Filed as one row because both halves are §Nouns of one page and neither is worth a verdict of its own; #75 is why two verdicts are not scoped to one sentence here. Depends on #120: that row removes eight duplicate entries, taking the Gaillean total from 512 to 504 — so the "Of 512 Gaillean entries" line on Dictionary:About Middle Airananwe and the class counts here should be re-derived in the same pass, not before it. Recorded as not a finding: everything else on the page checks out — the sixteen digraphs, the single-letter values and the five regular paradigms were read against Module:Auphen/YAQ-GA and Module:Gaillean and agree throughout. Update 2026-08-19 — the dependency is discharged, and §Nouns is now further out than when this was filed. #120 and #121 were carried out this morning, so the counts can be re-derived — and four of the six cells are stale, not one, because the figures the page publishes were the duplicate-inflated head-call counts. Against the entries as they now stand, one per page, and identical to what all sixteen Gaillean categories report, the table should read F 93, M1 71, M2 31, N1 44, N2 24, irregular 41: only M2 and the irregular total are already right. The entry total quoted on Dictionary:About Middle Airananwe as "Of 512 Gaillean entries" becomes 501 — 500 pages, ieca carrying two entries. (a)'s N2 half is answered rather than corrected: see #129. There are 25 class=N2 head calls in Gaillean and the twenty-fifth sits on a verb, so 24 is the true noun count, the appendix cell should read 24, and no -u noun is missing — the alternative this row offered can be dropped. (b) has moved too: "Their tables are not yet built" is now false of all three irregular classes rather than one, the feminine and neuter stems having landed this morning under #125, so the sentence wants replacing outright rather than narrowing. Re-derived 2026-08-20 and unchanged: counted over all 501 Gaillean headword calls, the six cells should read F 93, M1 71, M2 31, N1 44, N2 24, irregular 41, and all three irregular classes now have tables. #132 is this same question asked of Dictionary:About Middle Airananwe, which cites the parent's class count and gets a third figure again; answering both together keeps the two pages in step.
125	2026-08-18	structural	medium	21 Dictionary: entries (Gaillean, classes FX and NX)	Twenty-one Gaillean nouns are declared irregular and have no paradigm to be irregular against. Every one of the 283 regular-class noun entries carries a declension table, and since this window's last edit so do all 20 irregular masculines — Module:Gaillean holds their stems and the tables render. The 15 irregular feminines — aezhizel, carrix, elfitis, ennerd, gwani, hwith, ithlinexi-hwith, munu, nanti, qoethae, rian, sona, velna-hwith, wuerit, zwathor — and the 6 irregular neuters — allan, awiter, hipilas, ip, khest, mpes — carry class=FX or class=NX in the headword line, are categorised on it, and carry no table at all. The entries are not broken and no error renders: they simply stop where the other 303 nouns continue. The gap is structural rather than accidental — {{YAQ-GA-decl}} answers "Gaillean irregular noun: no stem recorded" when a stem is missing, so a table could not be added even if one were wanted. Dictionary:About Middle Airananwe already reads through the gap, mapping FX to Middle F1 and NX to Middle M1 as though the forms were settled.	flag-only — the fix is 21 stems, and a stem that "is not predictable from the citation form and must be given for each noun" (Dictionary:About Gaillean) is a creative act, not a maintenance one. Once the operator supplies them, the mechanical half is the same shape as this window's masculine pass: add them to Module:Gaillean, then add a Declension heading and {{YAQ-GA-decl|class=FX}} (or NX) to each of the 21 entries, which the routine can do in one pass. The endings themselves are already written down — Dictionary:About Gaillean gives -un, -i, -as, -i in the singular and -er, -tri, -z, -ter, -ia in the plural for the irregular classes — so it is the 21 stems and nothing else that is missing.	pending	New 2026-08-18. Not a defect report: the masculine irregulars were supplied the same day, so this reads as work in progress rather than an oversight, and it is filed so the remaining two thirds are not lost when the window closes. Related to #124(b), the sentence on the appendix page that says so; that half can be carried out now and this row left open. Recorded as not a finding beside it: 7 Gaillean noun entries carry no class at all — Inaea (Sun), Thaemo (Moon), axtat, iahaxtat, sponwa, tonhwith, velna-ollu — two of which are capitalised proper nouns for which no class may be deliberate; the other five are a smaller question of the same kind and are left inside this row rather than given one. Update 2026-08-19 — carried out. Module:Gaillean gained the fifteen feminine and six neuter stems this morning (rev 7438, the operator's, with the two derived compounds taking their head noun's stem), and all 21 entries gained a Declension heading and a {{YAQ-GA-decl}} call of the class their headword line names (revs 7439–7459). Verified rather than assumed: the module's stem table and the wiki's irregular entries reconcile exactly and in both directions — 41 stems, 41 entries in classes MX, FX and NX, no stem without an entry and no entry without a stem — and all 21 pages were parsed live, each rendering a table, with no "no stem recorded" fallback and no script error anywhere. Left pending for the verdict that clears it. The seven unclassed nouns recorded in these notes are unchanged, and today's pass sharpens two of them: velna-ollu stands beside ollu and ithlinexi-ollu, both MX and both now carrying stems, and tonhwith beside hwith, ithlinexi-hwith and velna-hwith, all FX and all now carrying stems — so those two kinship compounds read as the remainder of this pass rather than as a separate question. Of the other five, axtat, iahaxtat and sponwa have endings that Dictionary:About Gaillean's own rule reads off (-at, -at, -a), and Inaea and Thaemo remain the capitalised proper nouns. All seven titles are written unlinked, as before: five of them are on Special:LonelyPages.
126	2026-08-18	markup	medium	500 Dictionary: entries (Gaillean), Module:Auphen, Module talk:Auphen	The Dictionary's pronunciation engine cannot run the language added this window, and this routine's Auphen check cannot see that. Every Gaillean entry writes its pronunciation as a bare bullet — * /iət͡sʰa:/ — where Ancient Hertic (212 sections) and Besoh (1,096) both write * {{auphen|word|code}}, without exception. The reason is on the record: Module talk:Auphen carries a report filed this window that six of fourteen Gaillean test words return "Lua error: Internal error: The interpreter exited with status 1" — a hard interpreter exit, traced to a rule whose left-hand contexts run past the start of the word, and reproduced exactly (every word still holding ɛ when that rule is reached crashes; every word without one is correct). Module:Auphen/YAQ-GA itself is sound: it reproduces the ruleset on every word the engine survives. The check that should have caught this reads clean. Category:Auphen errors holds 0 and has held 0 all week, because an interpreter exit is not a tracked error — and because no live page calls the engine on Gaillean at all, the 500 entries carrying static IPA generated off-wiki instead. Nothing on the wiki is currently broken by it: 95 pages were rendered end to end this run, including all 20 irregular-masculine entries, and not one produced a script error.	flag-only as to the engine — Module talk:Auphen leaves that fix to Khurouan, which is right; this row exists so the operator sees it on the dashboard and not only on a module talk page. Two mechanical consequences follow and can be scheduled: (a) once the guard is in, convert the 500 Pronunciation lines to * {{auphen|<word>|YAQ-GA}} — a one-line pass per page, exactly as was done for Besoh, and the routine can carry it out; (b) until then, say somewhere a reader can see it that the Gaillean pronunciations are hand-generated rather than derived, since Dictionary:About Gaillean currently says only "Pronunciations are given as estimated IPA".	pending	New 2026-08-18. Not a criticism of the entries or of the data page: the static IPA is verified against the operator's ruleset word for word, and the language was added the only way it could be while the engine is out. Recorded as a limit of this routine: the Auphen errors check reads a category, so it reports what the engine says went wrong and is blind to what makes it exit — an engine fault of this kind will keep reading as 0 until it is looked for by hand. The same holds for script errors generally: this wiki has no Pages with script errors tracking category, so rendered errors are found only by parsing pages, which this run did for 95 of 2,622. Carried out in full 2026-08-20, and verified: Module:Auphen rev 7477 replaced the matcher that could not run Gaillean, and revs 7479 to 7978 converted all 500 entries to {{auphen}}. This run checked the result rather than taking it: the 500 generated pronunciations are identical, character for character, to the 500 static IPA strings they replaced, and running the old and the new engine side by side over all 1,816 distinct words on the wiki gives zero divergences, so no other language's pronunciation moved either. The row needs only a verdict to clear. What remains of it is filed separately as #135 — the live acceptance suite still tests one language of four.
127	2026-08-18	maintenance	low	28 Dictionary: entries (provisional Middle Airananwe)	The deletion queue grew by 28 in a single window, and all 28 are a language that was withdrawn hours after it was written. 55 Middle Airananwe entries were created this window from the new Gaillean lexicon, and all 55 were withdrawn the same day: 27 were sections on pages shared with another language and were simply removed, and the remaining 28 were whole pages, blanked and tagged. Every one carries the same reason, word for word — {{delete|Provisional Middle Airananwe entry, withdrawn pending the language's redesign}} — and every one is now nothing but that tag. The titles: Dictionary:-êue, abharizê, aghamu, aghessê, aguinsa, ala, alan, alantra, ankêue, annaue, annauzhtê, annauztrole, aqounê, arhirlê, artus, athlimbê, athuê, auarê, aubhientus, auiter, authat, avalius, bel, bezwetus, bhenkê, bhientus, bhinuê, bhirê. The withdrawal itself is clean: no Gaillean entry is left pointing at a removed form, no Descendants section anywhere on the wiki is empty, Lexemes holds zero Middle Airananwe rows, and nothing outside the tags and the two appendix pages mentions the language's forms. Category:Pages flagged for deletion now stands at 122.	flag-only — only an administrator can act on it. The editing bots hold no delete right, so this row puts the 28 in front of the one account that can remove them, and should be dismissed the way #58, #70 and #81 were, so that later runs list them under Deletion queue without re-flagging.	pending	New 2026-08-18. Not a finding against the withdrawal, which was done properly and completely. Recorded and not flagged: Dictionary:About Middle Airananwe now documents in full — nine sound developments, an orthography, a gender table — a language with no entries on the wiki, while Module:Head carries its noun classes and Module:Languages its code. That is documentation standing ahead of content, which is a reasonable place to leave a language mid-redesign, and it is worth watching only if the redesign moves the sound laws the page states. #119 and #123 are both on that page and are separate rows.
128	2026-08-19	structural	medium	Dictionary:mum	The scratch half of a Proto-South-Andusian entry lost its descendant when the hail half gained the right one. Dictionary:mum's Proto-South-Andusian section carries two numbered etymologies — Etymology 1, the verb scratch, incise, and Etymology 2, the noun hail — and one Descendants list between them, standing under Etymology 2. Until this morning that list read * {{desc|YCQ-BE|mum}}, which answered the Besoh verb mum, whose etymology opens "Inherited from Proto-South-Andusian {{m|YCP-XB|mum}}". #84 was carried out exactly as written and the bullet now reads * {{desc|YCQ-BE|mium}} — right for hail, and the section now names no descendant for scratch at all. The only Besoh word that claims this page as its parent is therefore unanswered by it. 1,484 explicit ancestor claims were matched against their parents' Descendants lists this run, in both directions: this is the one failure that is neither #87's twelve affix entries nor a row already on the table.	One line, and it is the line the page carried this morning. Under Etymology 1, beneath the verb's sense lines, add a Descendants heading and * {{desc|YCQ-BE|mum}}. Nothing new is asserted: Besoh mum already names this etymon, in this sense, on its own page, and the bullet is the one the section lost rather than one the routine has invented.	pending	New 2026-08-19, and created by carrying out a verdict: the edit #84 asked for was right and was made correctly. The fault is in that row's own parenthetical — "(the scratch section's descendant is correct and stays)" — which assumed a second Descendants list under Etymology 1. The page has never had one: it carried a single list, filed under the wrong etymology, and repointing it at the right word left the other etymology bare. Not a re-opening of #84, and not #112 or #113: those two turn on whether Besoh mum still holds the noun hail and on the five form entries standing beneath it, and this bullet is wanted whichever way they are read — the Besoh verb scratch is the sense neither row disputes.
129	2026-08-19	structural	low	Dictionary:wunei (title unlinked, see below), Dictionary:τυχασε	Two headwords carry an inflection class their part of speech has no paradigm for, and the parameter is accepted and discarded. Of the 448 {{head}} calls on the wiki that pass a class, 446 are nouns taking a declension or Ancient Hertic verbs taking a conjugation, and every one of them shows the class where a reader can see it — "aghammu n (second-declension)", "ανχοιαν m (first-declension)", "αργαετω (OO-class)". The remaining two show nothing. Dictionary:wunei is a Gaillean verb glossed bury and carries class=N2, a noun declension; Dictionary:τυχασε is an Ancient Hertic adjective glossed full and carries class=N1, likewise. Neither renders a class in its headword line, neither joins a class category, and no maintenance page can see either: both were parsed live to confirm it. The Gaillean one is the whole of #124(a). That row reports Dictionary:About Gaillean's N2 cell as 25 against 24 entries; there are exactly 25 class=N2 head calls in Gaillean, and 24 of them are nouns. wunei is the twenty-fifth.	Two words. Remove |class=N2 from wunei's headword line and |class=N1 from Dictionary:τυχασε's. Nothing rendered changes, no category changes, and the class-bearing head count then agrees with the class categories exactly, in both languages. If instead the classes are meant — Ancient Hertic adjectives agreeing in a noun class, or wunei being a noun its gloss does not describe — then the fix is not this one and is not maintenance: Module:Head would have to render what it currently drops, and that is a decision about the languages, not about the wiki.	pending	New 2026-08-19, from reconciling every class value against the part of speech it sits on for the first time. Each is the only one of its kind in its language: 82 Gaillean verbs, and one carries a declension class; 34 Ancient Hertic adjectives, and one carries one. It answers a question #124 could not — that row could say only that the wiki was "one -u noun short" of the figure the appendix publishes, because a count of head calls cannot see that one of the twenty-five is a verb; no -u noun is missing. Recorded as not a finding beside it: 56 Ancient Hertic verbs carry class=IS or class=OO and are right to — that language conjugates by class, the label renders, and the two wanted templates for those classes are already on the table as deferred #21. wunei's title is written unlinked here: it is one of the pages on Special:LonelyPages, and this page should not be its first incoming link (see #82).
130	2026-08-19	structural	low	Dictionary:so'-	The one Besoh affix that skips the step its five siblings all name. Six Besoh affixes are inherited from Proto-Andusian, and five of them write the chain out in full — "Inherited from Proto-Andusian {{m|YCP-XX|-me}}, through Proto-South-Andusian -me" — naming the middle form in plain italics precisely because the Proto-South-Andusian affix entries do not exist to link (#87). Dictionary:so'- writes "Inherited from Proto-Andusian {{m|YCP-XX|sog-}}; the same eroded word as the negator so'" and stops there. The step is not unknown: Dictionary:sog-'s own Descendants list gives the South Andusian reflex as sog-. A second, smaller mismatch on the same page: the entry's definition line ends with the worked example "(so'- + gob > {{m|YCQ-BE|soǵob}})", writing the element as gob, while Dictionary:saźgob and Dictionary:soǵob — both written or rewritten this morning under the ruling that made gob a bound element — write it -gob, with the hyphen that says so.	One edit, both halves. Complete the etymology to "Inherited from Proto-Andusian {{m|YCP-XX|sog-}}, through Proto-South-Andusian sog-; the same eroded word as the negator so'", the shape the five siblings use, with the middle form read off Dictionary:sog-. And write the worked example's element as -gob, matching the two entries that cite it. Nothing is invented and nothing is linked that does not exist.	pending	New 2026-08-19, and filed because the row that recorded it is being cleared: the omission was written into #89's Notes as something "recorded beside it", together with the sentence that answers it — Dictionary:sog-'s Descendants list gives the step — and #89 is resolved by the operator's fixed verdict this run, which would take the observation with it. Not a re-opening of #87: that row is about thirty-six branch-proto affix entries that do not exist, and this one asserts nothing about them — the middle form stays in plain italics, unlinked, exactly as it stands on the five siblings.
131	2026-08-20	structural	medium	Module:Head, Dictionary:About Middle Airananwe	The class registry still holds the gendered Middle Airananwe that was withdrawn. Module:Head keeps a CLASSES table naming the inflection classes each language admits, and its YAR-AI block declares four — M1 and M2, masculine, first and second declension, and F1 and F2, feminine — written on 2026-08-18 for the first Middle Airananwe design. The design ratified this window says the opposite twice: Dictionary:About Middle Airananwe §Entry conventions has "There is one declension; there is no gender", and §Nouns opens "One declension, no gender, four cases in two numbers." Nothing reaches the four: there is not one YAR-AI headword call anywhere on the wiki, the language's 28 provisional entries being queued for deletion (#127). The block is dormant rather than wrong today — but the first entry written against it would be labelled "m (first-declension)" and filed into a category the appendix says cannot exist.	Delete the YAR-AI block from CLASSES — eight lines at the head of Module:Head, no other language touched. Nothing rendered changes, because nothing calls it. Module:Languages's own YAR-AI entry, which supplies the language's name and section heading, is correct and stays.	pending	New 2026-08-20, from reading the machinery written for the language this window against the design ratified for it the same morning. Not #129: that row is about two live headword calls passing a class their part of speech has no paradigm for, and both calls exist; this is a registry entry no call reaches. Recorded as not a finding beside it: Template:YAR-AI-decl joined Special:UnusedTemplates this window for the same reason — the language has no entries yet — and that is work in progress, not a defect.
132	2026-08-20	factual	low	Dictionary:About Middle Airananwe, Dictionary:About Gaillean	One page says nine, the other says six, and the entries use eight. Dictionary:About Middle Airananwe §Nouns closes on the parent language: "Gaillean's five cases lose the vocative here; its three genders and nine stem-classes are gone altogether." The five cases and the three genders check out against the parent page. The nine matches nothing: Dictionary:About Gaillean §Nouns opens "Gaillean nouns fall into six declensions" and tabulates six rows — F, M1, M2, N1, N2, and one row bundling the irregular MX, FX and NX — while the entries themselves use eight class codes, counted over all 501 Gaillean headword calls: F 93, M1 71, M2 31, N1 44, N2 24, MX 20, FX 15, NX 6. Nine is what the two records give added together: the table's six rows, plus the three irregular classes its sixth row already bundles.	One word. Either six, matching the parent page's own count, or eight, matching the codes the Dictionary files — and the same choice is already before the operator as #124(a), which asks what Dictionary:About Gaillean's own §Nouns should say. Answering both the same way keeps the two pages in step.	pending	New 2026-08-20, from checking a page rewritten this window against the page it summarises. Independent of #124: the figure is wrong on either reading of that row, so this can be settled first or together, but not by ignoring it. Recorded as not a finding: the rest of the sentence holds — the vocative is in the Gaillean paradigm table and absent from the Middle Airananwe one, and the three genders are the masculine, feminine and neuter of the parent's own table.
133	2026-08-20	structural	low	Dictionary:About Middle Airananwe, Module:MiddleAirananwe	The generated table drops a letter the documented paradigm keeps. Module:MiddleAirananwe, written this window to back the noun table, discards a word-final -e before the dative singular -i — added at the operator's instruction in rev 7476, whose summary gives the case as ghale to ghali. Dictionary:About Middle Airananwe §Nouns sets out the paradigm in full — four cases, two numbers, "the ending attaches to the bare lexeme, taking a linking -e- after a consonant; the dative singular needs none, being vocalic itself" — and says nothing about the loss. The two disagree in the reader's face on every noun in -e: the table shows ghali, the documented paradigm predicts ghalei. It is the only rule of the declension that the appendix does not carry.	One sentence, restating what the module already does and asserting nothing further. After "the dative singular needs none, being vocalic itself", add: "a word-final -e is discarded before it", with the module's own worked pair ghale / ghali as the example. The wording is the operator's; the rule is not, being settled in rev 7476.	pending	New 2026-08-20, from reading the new module against the appendix page it backs. No entry is affected today: the language has no live entries, so nothing renders the discrepancy yet, which is why it is filed low and why it is cheapest now. Recorded as not a finding beside it: the rest of the module reproduces the appendix table cell for cell, including the dative's lack of a linking vowel after a consonant, so ghal and ghale both give ghali — a syncretism the paradigm implies and the module carries out faithfully.
134	2026-08-20	markup	low	Dictionary:About Gaillean, Dictionary:About Middle Airananwe	The broken helper of #79 is on three pages now, and the newest was written this window. A {{tl|…}} call renders a red link, because the wiki has no such template (title written unlinked — see Notes). #79 names the two calls on Dictionary:About Besoh and proposes the fix. Two more are live and outside its scope: Dictionary:About Gaillean §Entry conventions — "Nouns carry their class in the headword line and a declension table ({{tl|YAQ-GA-decl}})", written 2026-08-18 — and Dictionary:About Middle Airananwe §Entry conventions — "Nouns carry a declension table ({{tl|YAR-AI-decl}})", written this window at rev 7475. Four calls on three in-frame pages, one missing template; parsing each page shows the red link.	Whatever settles #79, applied to all three pages in the same pass — under that row's (a), the table names go in plain italics or <code> tags, which is what the other three Dictionary:About … pages do; under its (b), the missing template is created once and all four calls come good together. This row asks only that the verdict be read as covering the two later pages, not that a different fix be chosen.	pending	New 2026-08-20. Filed rather than folded into #79: that row's Suggested fix names one page and its own quotations are from that page, and #75 is the standing example of what happens when one verdict is assumed to reach text it does not name. The template's title is written unlinked throughout this row, per #117: writing it as a link would emit the red link the row is about, from this Report. Recorded as not a finding: the escaped mentions in #79 and here are inside <nowiki> and do not execute.
135	2026-08-20	structural	low	Template:Auphen/testcases, Module:Auphen	The pronunciation engine was rewritten this window and its live test suite exercises one language of four. Template:Auphen/testcases is the wiki's own acceptance test: each row compares the engine's output to a hand-checked expectation and files the page into the failure category when they part. It holds four rows, all Pjany — the language with 7 calls on the wiki — plus one graceful-failure case for an unknown code. Besoh (1,100 calls), Gaillean (501) and Ancient Hertic (212) have no row at all, so a drift in any of them files nothing and shows nowhere. The engine's matcher was replaced wholesale at rev 7477, which is exactly the change such a suite exists to police.	Three rows, in the shape the page already uses — one word per uncovered language, with the expectation taken from a value the wiki already vouches for rather than from the engine's own output. Gaillean is the cheapest: any of the 500 static pronunciations replaced this window, verified against the operator's own ruleset before they were replaced, e.g. artuz for äɹtus. A Besoh and an Ancient Hertic word chosen the same way complete the set. Which words is a small editorial choice; the shape of the rows is mechanical.	pending	New 2026-08-20, from checking the window's engine change rather than from the suite itself, which passes. The gap is not hypothetical and not open: this run reran every one of the 1,816 distinct words the wiki asks the engine for through both the old and the new code and found the outputs identical, and separately checked all 500 new Gaillean calls against the static IPA they replaced — also identical, character for character. That check lives in this report, not on the wiki, which is the whole of the row. Distinct from #126, which is about the engine being unable to run Gaillean at all.
136	2026-08-20	voice	low	Dictionary:About Middle Airananwe	Two sentences on a page rewritten this window argue for the subject rather than describing it. Yezur Wiki:Conventions asks for an encyclopedic register — "Neutral and third-person, definition-first … No storytelling, no chatty asides … no marketing or how-to tone." Two lines step outside it. §Phonology: "The palatal series … is the language's signature, and the source of the palatals of the modern language." §Development from Gaillean: "The ending system is the reason the stage matters." Both state the importance of the topic to the reader instead of stating a fact about it, and both are carrying claims that stand on their own — the palatal series is the source of the modern palatals; the ending system distinguishes the three stages, as the four sentences after it show.	Two clauses, no claim lost. "…is its most distinctive feature, and the source of the palatals of the modern language", or simply "…is the source of the palatals of the modern language". And for the second, "The ending system separates the three stages", which is what the sentences that follow it demonstrate.	pending	New 2026-08-20, from the conventions read of the window. Filed low and easily rejected: the appendix pages are practice documentation rather than articles, and emphatic lead-ins are house style there — Dictionary:About Besoh has 28 bold runs and Dictionary:About Proto-Andusian 10, and this row is not about those. What is flagged is the evaluative claim, not the emphasis. Recorded and not flagged: "Middle Airananwe is not a mechanical halfway point between the two attested stages" reads as an ordinary statement about a thinly attested intermediate stage, and is left alone.

Dismissed flags

Deferred or rejected findings, kept here so later runs do not re-flag them.

• #8 (deferred) — Wanted history-period pages Yezuri Age of Exploration / Yezuri Middle Ages. Operator: not canon names for these eras; will not create stubs under temporary names.
• #11 (deferred) — Andusian / Tacchoan / Shung-Worrehan member-language queries fall back to the default. A content gap, not a data error; operator: no member languages are in yet, add when ready.
• #12 (rejected) — Greek-letter native names on Hertia / Hertic language / Alompe etc. Operator: fully intentional in-world orthographic style.
• #14 (deferred) — Uncategorised portal front pages and the root Category:Yezur. Operator: accepted, but no decided approach for handling it yet.
• #21 (deferred) — The four wanted Ancient Hertic inflection templates (conj-IS, conj-OO, decl-N1, decl-N2). Operator: "will be fixed when the tables are created — already on my to-do." Confirmed present on the to-do list at User talk:Yezur.
• #30 (rejected) — The one-line User:Yezur page, read as an unfinished sentence. Operator: the line is not unfinished — the city name is written there as , "using a script I have added to the wiki, with characters stored in the PUA … a temporary text intended to be used for testing the PUA font; it is working correctly." Answering the operator's condition for reflagging: the routine can see the characters and reads five PUA codepoints there (U+E067 U+E111 U+E179 U+E179 U+E0EB, the third and fourth identical, so a five-letter name with a doubled letter), which round-trip intact through the API. What it cannot do is read them: no mapping from those codepoints to sounds or letters exists anywhere on the wiki, so the glyphs are legible only to a reader with the font. Not reflagged. If a transliteration table for the script ever lands on-wiki, say so and later runs can check such names instead of skipping them.
• #32 (rejected) — Airananwe and the code name:yas-ai in Map:name's worked example matching no language and no registry code. Operator: "that language will land later. No need right now." Left as written; the routine will not re-open it when the language arrives.
• #33 (deferred) — The two redlinked values Map:place=hamlet and Map:place=suburb in Map:place's settlement table. Real but not now; they remain the only redlinked values in any key page's value table, and both are listed in the new Map:Features index only through their parent key.
• #43 (rejected) — Template:Infobox former country declared and stored into a FormerCountries Cargo table that was never created, so Hertia's structured data went nowhere. Operator: "Cargo tables no longer the go-to." Overtaken by the retirement of Cargo from the encyclopedia (2026-07-30): the declaration and the store have both been removed from the template, so there is no longer a table to create. Confirmed this run — of the six tables the wiki once declared, only Lexemes answers a query; Continents, Languages, LanguageFamilies, Polities and FormerCountries now all throw the same MWException that once distinguished an absent table from an empty one, because all five are now absent. The administrative Create data click this row was waiting on is no longer wanted.
• #59 (rejected) — Betonia, Unteria and Sea of Sihia categorised but reaching no Cargo table, with no infobox on the wiki fitting a subcontinent, an island or a sea. Operator: "No longer relevant since Cargo tables no longer the go-to. Not everything needs to be tabulized." The structured-data half of the finding has been removed by the same ruling, and the infobox question closes with it. Vetsihia's borrowed {{Infobox continent}} is left as it stands: with Cargo gone it is display only, and its type=Geographic region (colloquial) already says what the article is.
• #47 (deferred) — Meta: and Help: documented as live namespaces while holding no page, and the sidebar's Project docs link leading to an empty index. Operator: "scarcely used, but will become hugely useful should the project ever open to others — there is simply nothing yet to meta about on a large scale," and Help: is in the same position; no edits to MediaWiki:Sidebar, Yezur Wiki:Main Page or Yezur Wiki:Conventions. The amendment's own task — check whether any Yezur Wiki: page could move to Meta: — was carried out this run, and its one finding is filed as #73. Standing consequence, recorded rather than re-flagged: the sidebar link every reader sees still leads to an empty index, by decision.
• #50 (rejected) — The unreachable index-only branch in Engine:handle_affix of Module:Auphen, where the ELS ~= guard makes the second branch untestable. Operator: "let's not mess with the module's code." The routine will not re-open it; if an affix rule with an index-only else clause is ever written and silently does nothing, that is this row.
• #53 (deferred) — The OpenStreetMap attribution strip baked into File:Tacchoan-subfamilies.png, and the colour key that gives yellow to two branches. Operator: "the operator will upload a better export in time." The licence question is closed: the map is not OSM-derived — it only uses the OSM software — so no credit is required and the strip can simply go. Part (c), the caption clause, was carried out under #55 this run.
• #57 (deferred) — The four Tacchoan branch names (Esselis, Rumesselis, Kipoa, Usy) and the island Usings named into existence with no article, and linked in the caption but plain in the bullets that define them. Operator: deferred. The five remain the only red links out of Tacchoan; the branch names stay plain in the bullets.
• #58 (deferred) — The five Talk:Family A–Talk:Family E pages queued for deletion since 2026-07-28. Operator: "the operator will get to the deletions in due time." Listed under Deletion queue each run as the task list requires, but not re-flagged.
• #65 (deferred) — Map:Conventions, promised by Map:Main Page's lead and linked from two pages, never written. Operator: deferred. It remains the only red link in the Map: namespace outside dismissed #33.
• #70 (deferred) — The 37 superseded Proto-Andusian entries at implosive titles, queued for deletion by the operator's own ruling of 2026-08-03. Operator: "as #58, the operator will get to the deletions in due time." Listed under Deletion queue each run, not re-flagged.

Latest run

2026-08-20. No verdicts were waiting, so the routine changed nothing. All 54 rows stood at pending when the run began, so phase 1 had nothing to carry out and this Report is again the only page it has written. The window since the last run holds 507 changes — rcid 7450 to 7956 — of which one is the last run's own post and 506 are Thiorosan's, all made between 08:17 and 08:41 on 2026-08-20. Five of those are substantive; the other 501 are one mechanical pass and one follow-up.

What the window did

• The pronunciation engine was repaired. Module:Auphen rev 7477 replaced the matcher that enumerated category cross-products with one that tests membership per position — the "tracked follow-up" its own talk page had been carrying since the engine was written. Module talk:Auphen rev 7478 corrects the earlier diagnosis, which had read the crash as an out-of-range index.
• Gaillean joined the engine. All 500 Gaillean entries — revs 7479 to 7978 — exchanged their static IPA bullet for a {{auphen|word|YAQ-GA}} call. This is the pass #126 describes, carried out in full.
• Middle Airananwe was rebuilt to the ratified design. Dictionary:About Middle Airananwe rev 7475 is a rewrite — its own phonology, a palatal series, prenasalised stops, written vowel length, fifteen sound correspondences from Gaillean, and one genderless declension. Module:MiddleAirananwe and Template:YAR-AI-decl are new, and back that declension.

The engine change was checked, not assumed

A rewritten matcher under 1,825 live calls is the one thing in this window that could have gone wrong quietly, so it was tested rather than read:

• Every word, both engines. Module:Auphen at rev 185 and at rev 7477 were run side by side over all 1,816 distinct words the wiki asks the engine for, in all four languages. Zero divergences.
• Every new Gaillean call, against what it replaced. The 500 generated pronunciations were compared to the 500 static IPA strings edited out from under them — the strings verified against the operator's own ruleset word for word before the conversion. All 500 match, character for character.
• The engine's own suites. Template:Auphen/testcases passes, and Category:Auphen test failures and Category:Auphen errors are both empty. The 340-case acceptance suite the engine was built against passes on both revisions.
• The fix does what it says. On Gaillean input the old matcher takes about 82 ms a word and the new one about 0.4 ms — the two-hundredfold gap that was overrunning the interpreter and returning a crash rather than an answer.
• Filed from this: #135, because the wiki's own live suite tests Pjany only, and none of the above is visible on the wiki.

The findings

Six, all from the five substantive pages:

• #131 (medium) — Module:Head still registers four gendered Middle Airananwe noun classes from the design that was withdrawn; the appendix ratified this window says the language has one declension and no gender. Nothing calls them yet, which is why the fix is free today.
• #132 — the same appendix credits Gaillean with nine stem-classes; its parent page says six and the entries use eight.
• #133 — Module:MiddleAirananwe drops a word-final -e before the dative singular; the paradigm the appendix prints does not mention it, and predicts a different form.
• #134 — the missing-template call of #79 is now on three pages, the newest written this window.
• #135 — the live acceptance suite for the pronunciation engine covers one language of four.
• #136 (voice) — two sentences of the rewritten appendix argue for the subject rather than describing it.

Six standing rows were re-checked and their Notes updated

• #119 (voice, high) — resolved by the window. The rewrite deleted the §Gender section outright, and with it the four-word comparison the row is about. The wiki now holds no in-frame use of any of the 35 real-world terms swept for. The row needs only a verdict.
• #126 — carried out in full, by the two changes above. It too needs only a verdict.
• #107 — re-counted: the entries putting Pronunciation before Etymology now number 807 (Besoh 307, Gaillean 500) against 213 that follow the skeleton. The row was filed when the split was 306 to 213, and its (a) is now four times the work it was.
• #123 — stands. The rewrite kept the claim in new words: "No native name is recorded for it." Both rosters still give Airannawê.
• #124 — re-derived from the live entries and unchanged: F 93, M1 71, M2 31, N1 44, N2 24, irregular 41. #132 is the same question asked of the page that cites it.
• #79 — see #134: two later pages carry the same call.

What came back clean

• Structured data. Lexemes holds 3,239 rows and the wiki makes 3,239 headword calls — equal, and equal language by language across all eleven codes. No row is missing a key field, no row points at a page that does not exist, and every row's word matches its page title. The 797 empty glosses are Besoh form entries, which is correct; the five repeated signatures are the known homographs, each with distinct senses.
• Markup and machinery. Linter errors 0. Pages with script errors 0, with template errors 0. Auphen errors 0. Double redirects 0, broken redirects 0. Unused and uncategorised images 0.
• Hand-maintained lists. Each family's Members table still matches the family field of its languages' infoboxes exactly — Herta-Olaic 2, Njeshan 7, Andusian 1 — and Main Page links Category:Continents rather than enumerating it, so there is no list left to drift. The fifteen Gaillean source words cited by the rewritten appendix all exist, and all carry a Gaillean section.
• Counts, all unchanged unless noted. Wanted categories 125; wanted pages 62 (up one: Module:MiddleAirananwe/doc, the thirteenth Scribunto documentation placeholder, which belongs with the module-documentation item on the to-do list); wanted templates 5; uncategorised pages 10; uncategorised categories 1; uncategorised templates 3 (#98); unused templates 4 (up one: Template:YAR-AI-decl, which has no entries to sit on yet); dead-end pages 381.
• The deletion queue holds 122 pages, the same 122 as last run, page for page. #81, #82 and #127 still put them in front of the one account that can act.
• Orphans. Special:LonelyPages reads 504; four of those pages are masked by this Report's own links — Dictionary:About Besoh and the Gaillean entries bell, beule and peramme — so the true count is 508, unchanged. This post masks the same four, so the special page will read 504 again.

What is pending

60 rows, #72 to #136. #82 remains the only high row of substance; #119 is the other high and is now spent. Five rows need nothing but a verdict, their work being done or moot: #119, #120, #121, #125 and #126. Fourteen more need nothing but a yes: #98, #101, #102, #104, #105, #106, #109, #111, #114, #115, #116, #117, #122 and #129 (#116 waits on #108). Order matters in one place: #131 is free while the Middle Airananwe lexicon is empty and stops being free the moment an entry is written against it.