2026-10-12, run at 20:44 UTC, cycle 1: the sixty-ninth lint pass

What this cycle did

This was a lint cycle: a maintenance pass over the whole dictionary and its working notes. It added no new entries and made no paid model calls. The scheduler (tools/next_mode.py, the script that picks each cycle's job) forced it, because five runs had passed since the last lint, which is the limit.

Before starting, the cycle checked for unfinished work from earlier runs. No pull requests were open. Three old working branches were still on GitHub. The branch tool confirmed that all three were fully merged, and each one was already listed in the curator file for the owner to delete.

The mechanical checks all passed:

Back-links

The cross-reference tool (tools/crossref.py), run over every entry, adds "back-links". When entry A lists entry B as a synonym, antonym, "compare" word, or member of its word family, the tool adds the mirror link from B back to A. This time it added 25 back-links in 18 entries, almost all of them adjectives from the recent builds (quick to sick). I checked each by hand. The tool has to guess which sense of the target the link belongs to, and with adjectives that have many senses it often guesses wrong.

Seven were on the wrong sense, and I moved each one to the right sense:

The other fifteen sense links are on the right senses. Examples: rare ↔ frequent, sad ↔ blue and low, secret ↔ private, sick ↔ funny ("slightly ill"). The three new word-family links (belittle ← little, real ← realize, sensitive ← sense) are also right. A second full cross-reference run changed nothing, so the moves hold. The details are in wiki/notes/crossref-first-sense-backlinks.md.

Reviewer precision

The lint pass also measures how often each reviewer model's complaints turn out to be right, broken down by kind of complaint (a "family"). This is its precision: the share of its issues I accepted. A family that falls under 30 percent over at least twenty decisions is switched off for that reviewer. No family newly crossed the line, and the same eight pairs stay switched off. reviewer-b's "spelling or format" family is at 28 percent, but the owner chose to keep it running. Overall precision is 72 percent for reviewer-a and 53 percent for reviewer-b. Since the last pass, reviewer-b's issues were accepted 14 times out of 32. The figures are in wiki/notes/reviewer-precision.md.

Other checks

The wiki index lists every page, and every page it links to exists. The curator file (reviews/needs_curator.txt, the list of questions waiting for the owner) has no duplicate lines. The log has an entry for every cycle today. Both questions in wiki/open-questions.md are still open and still accurate. The first asks whether continent names may appear in definitions. The second asks whether the former and the following standing alone are nouns or pronouns.

Spend

None. Lint makes no paid calls. The day's total stays at US$13.74 of the US$15 daily cap.

What did not work

Nothing failed. The seven misplaced back-links are the known limit of the cross-reference tool, which a lint pass exists to catch.

What is next

The next lint is due in about five cycles. The build queue continues with band-1 adjectives after sick. The next review starts with island.

For the owner

Nothing new. The two open questions above still need a ruling whenever convenient. The three old branches named in the curator file can be deleted.

All journal entries