Journal — 2026-09-19 (run 4 of the day)
This was the sixth scheduled run of the "Routine," the unattended process that drafts, checks, and publishes entries for the TKG English Learner's Dictionary (a project made of one small JSON file per dictionary entry, published as a website). Nobody reads this in real time; it is the record for the owner (Tom) to catch up on later. The mode selector chose "build" again: write new dictionary entries.
Start-of-run housekeeping
No pull request was left open from a previous run, so there was nothing stranded to merge or fix. No orphan branches needed sweeping. The inbox/ folder — where the owner can drop instructions that jump the queue — held only its archive of already-handled items. The checkout was confirmed to hold the latest work from main.
The main work: twelve pronouns
I claimed and drafted twelve pronouns (words that stand in for a noun phrase on their own, without a noun directly after them): anyone, anything, both, each, either, enough, everybody, everyone, everything, few, he, her. Several of these continue the paired determiner/pronoun split from earlier runs — "both" before a noun (both hands) already has its own entry; this run added "both" standing alone (I'll take both).
For every entry I wrote the definitions, examples, and grammar notes, then ran the project's full quality pipeline: a form checker, a pronunciation checker (which compares my proposed pronunciations against three other AI models and an open pronunciation dictionary), and a two-reviewer panel of two more AI models that read every field and flag anything they consider wrong.
The panel flagged 45 issues, 30 of them serious enough to block publication. I read every one against the entry and my own knowledge before deciding; I applied 30 fixes and rejected 15 objections.
The most useful pattern in the fixes: five entries had let a grammar fact ("standing alone without a noun") leak into the definition field itself, when that fact belongs in the explanation instead — a definition should say only what the word means, not how it behaves grammatically. Once one reviewer caught this on "both," I found and fixed the same slip on "each," "either," "enough," and "few" myself. Separately, "each"'s own definition had accidentally used the word "each" to define "each" — a circularity the style guide forbids — and two of its example sentences ("the tickets cost ten dollars each") turned out to illustrate a different, adverb-like use of the word, not the pronoun sense this entry covers, so I removed them along with the matching example under "learner mistakes."
I rejected three pronunciation objections outright, because an independent check (agreement with a public pronunciation dictionary, done automatically before the reviewers ever see the entry) had already confirmed my transcriptions were right and the reviewer's proposed alternative was not standard. I also rejected two objections that would have contradicted claims already published, reviewed, and standing unchallenged in the sibling entries for "both" (as a determiner) and "few" (as a determiner) — changing this run's pronoun entries to disagree with those would have made the dictionary inconsistent with itself.
Checks, cost, and what's next
Every automated check passed: the schema validator (across all 132 entries now in the dictionary), the word-count limits on key project files, the internal-link checker, 292 automated tests, and the closed-vocabulary checks. Today's spending is $0.93 of the $5.00 daily budget; this run's share was $0.28.
Nothing new needs the owner's attention beyond what was already waiting (a few old branches this tool cannot delete, one disputed British pronunciation, and the three items already logged in NEXT.md). The next run is forced by a project rule to be an "originality" check — a periodic search of the web to make sure the dictionary's wording is not accidentally matching a published dictionary's — rather than whichever mode the usual scheduler would otherwise have picked.