D345 — kb_review ignores series pages and historical links

Decision. duplicate_candidate (title check) and the creation-time similar advice skip two pages that form a series: titles equal but for period tokens (a token with a digit, or q1-q4/h1-h4) and id basenames differing only in those positions (lint.SeriesSiblings). zombie_work counts only links under an origin/dependency heading (built-in English and Italian list), not links in prose.

Why. Both were false positives that operators silenced with lint_ignore, which also hides future true positives on the same page. Quarterly pages share most title words by construction; a precedent cited in prose is history, not unfinished work, and "close it as obsolete" is wrong advice for it.

Alternatives rejected. - A map-contract key for the origin headings: the map_update catalogue is about one byte from the D285 budget. - Keying the series on titles alone: two unrelated pages with numbers in the title would vanish from review; requiring the ids to follow the same pattern guards that. - Month names as period tokens: language-dependent, would need a vocabulary. - Frontmatter or declared link fields as dependency evidence: the link graph is body-only (D241) and no field carries concept links; a separate feature.

Consequences. One shared helper serves both call sites; change it with both tests. Origin links resolve through ExtractLinks with the concept's real file path (expanded concepts). A KB in another language uses lint_ignore: [zombie_work]. link_to_retired and stale_open still read body-wide links.