skills(running-tend): name the bundled gate this section overrides

Gate 3 in the bundled review-gates reference classes a never-restoring
cache as wasted compute and names "a cache" among the mechanisms that
fail it at any occurrence count, so the delete-and-comment call this
section corrects is what that gate prescribes. Framing the section as a
reading of the gate left the two disagreeing; framing it as a repo
override resolves it under the documented repo-over-bundled precedence.
This commit is contained in:
worktrunk-bot
2026-08-18 21:13:02 +00:00
parent 8ceba037ef
commit 86177c14e8
+5 -4
View File
@@ -118,10 +118,11 @@ restores, a step whose output nothing reads, a check that can't fail — should
open the PR that makes it work. Deleting the broken thing and leaving a
comment explaining what a working version would take is the wrong shape: it
banks the analysis in a code comment and leaves the improvement unclaimed,
and a maintainer reading it asks for the fix anyway (#3846). The
review-runs cost gate is about not spending a maintainer's review on churn,
not about preferring the smaller diff — a fix that is three coupled edits
with a real payoff clears it.
and a maintainer reading it asks for the fix anyway (#3846). This overrides
**Gate 3** in the bundled review-gates reference for this repo: that gate
classes a never-restoring cache as wasted compute and names "a cache" among
the mechanisms that fail it at any occurrence count. Here, machinery a sweep
has diagnosed gets the fix rather than the deletion the gate would allow.
Fix it when the mechanism is understood well enough to predict the working
version's behavior offline (recompute the key, replay the hash, diff the