From f7a4997bdcd8fa0296dacc431fb69c2000fb65a2 Mon Sep 17 00:00:00 2001 From: Aashil Date: Tue, 29 Sep 2026 22:24:39 +0545 Subject: [PATCH] chore: offset crisp-triage cron off the exact hour Two consecutive scheduled slots (12:00 and 15:00 UTC) never fired at all on 2026-09-29 -- no failed/skipped run logged, just silently absent. Workflow confirmed active, YAML valid, cron expression valid. Matches GitHub's own documented cause: every scheduled workflow across GitHub competing for the exact top of the hour is a known source of delayed/dropped runs. Not a proven root cause (no access to GitHub's internal scheduler), but a low-risk mitigation per their own guidance -- if the gap recurs on this new offset, that would actually rule this out and point elsewhere. Co-Authored-By: Claude Sonnet 5 --- .github/workflows/crisp-triage.yml | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/.github/workflows/crisp-triage.yml b/.github/workflows/crisp-triage.yml index fee8a00..2fface1 100644 --- a/.github/workflows/crisp-triage.yml +++ b/.github/workflows/crisp-triage.yml @@ -12,7 +12,12 @@ name: Crisp triage on: schedule: - - cron: "0 */3 * * *" + # Minute 17, not 0 -- every scheduled workflow on GitHub competing for the + # exact top of the hour is a documented cause of delayed/dropped runs. + # Confirmed for real: two consecutive 3h slots (12:00 and 15:00 UTC) + # never fired at all on 2026-09-29, no failed/skipped run logged either, + # they just silently didn't happen. + - cron: "17 */3 * * *" workflow_dispatch: inputs: skip_dedupe_check: