pr-review-conduct gates the dispatched-escalation route on the seat rather than on being dispatched:
An agent that cannot reach the maintainer directly, a dispatched subagent being the ordinary case, escalates to whoever dispatched it and stops that unit of work there.
drive-pr "Disposing of Every Finding" and local-strict-review "Disposing of Findings" now match that gate. The gate is wider than the remedy: an interactive seat whose maintainer is simply away meets "cannot reach the maintainer directly" and has no dispatcher to escalate to. pr-review-conduct covers that case inside outcome 3 itself ("If the maintainer is not reachable right now, leave the thread open and say so"), and neither of the two other surfaces carries it or routes to it.
It bites hardest in local-strict-review, where the pass is a mandatory pre-push gate: an interactive seat with an outcome-3 finding and an unreachable maintainer is told the answer is needed "in the same turn", is offered an escalation route it has no dispatcher for, and has nothing that clears the push.
Two smaller ones from the same sweep:
drive-pr's frontmatter description is the one of the three statements that drops the stop. It says "escalating to whoever dispatched the drive instead where the maintainer cannot be reached from that seat" where the body and pr-review-conduct both say the seat "stops that unit of work there". Escalate-then-keep-moving is the failure that section names explicitly. The same clause's "that seat" also has no antecedent in the description.local-strict-review's escalation clause says the escalation "stops the work there", where the source scopes the stop to "that unit of work" and adds "A dispatcher's other work is not stopped by it".
And one that the same rewrite put in tension with a neighbour: local-strict-review says "The findings stay advisory, and the count a pass raises gates nothing at all", while an outcome-3 escalation now stops the push. The unit already handles one such pair explicitly ("Two claims sit next to each other here and they point opposite ways"), and that treatment is about mandatory-pass versus advisory-findings and does not cover this third claim.
Raised by the local-strict-review carried-content passes on the pull request that added backlog-burndown, which edited these sections, and deferred there under that skill's own review-round budget.
pr-review-conductgates the dispatched-escalation route on the seat rather than on being dispatched:drive-pr"Disposing of Every Finding" andlocal-strict-review"Disposing of Findings" now match that gate. The gate is wider than the remedy: an interactive seat whose maintainer is simply away meets "cannot reach the maintainer directly" and has no dispatcher to escalate to.pr-review-conductcovers that case inside outcome 3 itself ("If the maintainer is not reachable right now, leave the thread open and say so"), and neither of the two other surfaces carries it or routes to it.It bites hardest in
local-strict-review, where the pass is a mandatory pre-push gate: an interactive seat with an outcome-3 finding and an unreachable maintainer is told the answer is needed "in the same turn", is offered an escalation route it has no dispatcher for, and has nothing that clears the push.Two smaller ones from the same sweep:
drive-pr's frontmatter description is the one of the three statements that drops the stop. It says "escalating to whoever dispatched the drive instead where the maintainer cannot be reached from that seat" where the body andpr-review-conductboth say the seat "stops that unit of work there". Escalate-then-keep-moving is the failure that section names explicitly. The same clause's "that seat" also has no antecedent in the description.local-strict-review's escalation clause says the escalation "stops the work there", where the source scopes the stop to "that unit of work" and adds "A dispatcher's other work is not stopped by it".And one that the same rewrite put in tension with a neighbour:
local-strict-reviewsays "The findings stay advisory, and the count a pass raises gates nothing at all", while an outcome-3 escalation now stops the push. The unit already handles one such pair explicitly ("Two claims sit next to each other here and they point opposite ways"), and that treatment is about mandatory-pass versus advisory-findings and does not cover this third claim.Raised by the local-strict-review carried-content passes on the pull request that added
backlog-burndown, which edited these sections, and deferred there under that skill's own review-round budget.