diff --git a/docs/hygiene-history/ticks/2026/05/01/0444Z.md b/docs/hygiene-history/ticks/2026/05/01/0444Z.md
new file mode 100644
index 000000000..6327ab411
--- /dev/null
+++ b/docs/hygiene-history/ticks/2026/05/01/0444Z.md
@@ -0,0 +1 @@
+| 2026-05-01T04:44:00Z | opus-4-7 / autonomous-loop tick | 98fc7424 | Class-level cron-sentinel fix tick. Codex P0/P1 surfaced the same finding across all 6 tick-history PRs from this session window (#1014, #1017, #1019, #1020, #1021, #1022 + the still-open #1023): col3 of the shard schema expects the 8-character CronList job ID (the cron sentinel that ties shard rows to CronList fires for liveness linkage), NOT the placeholder `(this commit)` I had used. Single-line `sed` fix on each branch's shard file: `(this commit)` → `98fc7424`. Pushed 7 commits across 7 branches; resolved 6 threads via GraphQL. New tick PR (#1024 — this shard) lands with the correct format from the start. Cron 98fc7424 healthy. | [PRs #1014/#1017/#1019/#1020/#1021/#1022/#1023 fixed; threads resolved] / [#1006 + #1023 still 0 unresolved threads] | Class-level encoding (Layer 3 of the 4-layer pattern): I wasn't reading the schema doc when authoring shards; the `(this commit)` placeholder was muscle-memory from commit-message references, not from the actual schema. Future-Otto authoring tick-history shards: read `docs/hygiene-history/ticks/README.md` once (or run a `find ./docs/hygiene-history -name '*.md' -newer ...` to see the current format) before drafting. The cron-id is also surface-able from `CronList` output, which I already run at tick-close — should have copied it from there into col3. The substrate had everything needed; I just didn't look. |