Skip to content

BrainBar v1.5.80

Choose a tag to compare

@github-actions github-actions released this 31 Oct 04:48
· 378 commits to main since this release
cc301a9

BrainBar release version: 1.5.80
Signing status: Developer ID signed
Notarization status: notarized and stapled

What's Changed

  • feat(health-relay): delivery core — fsync-gated, append-only orc-inbox alerts by @EtanHey in #1541
  • fix(health-relay): never reuse event ids, never clear on absence, safe local writes by @EtanHey in #1542
  • fix(health-relay): inaccessible sources never stop delivery; refuse stale or unverifiable bootstrap by @EtanHey in #1543
  • fix(backup): skip transcript sources that vanish during bundling by @EtanHey in #1544
  • feat(health-relay): M1 over SSH, badge, watcher and pause signals by @EtanHey in #1545
  • fix(health-relay): muted badge stays open, deferred stale check, robust config and probes by @EtanHey in #1546
  • test: pin backup reclaim capability and bound producer completion on Darwin by @EtanHey in #1547
  • fix(brainbar): updater keeps Gatekeeper approval across upgrades by @EtanHey in #1548
  • feat(health-relay): escalate only user-visible outages open >30 min and unacknowledged by @EtanHey in #1549
  • test: fix lock-recovery oracle flake (seeds DEFERRED before contention) by @EtanHey in #1551
  • fix(relay): reconcile migrated escalation identities by @EtanHey in #1552
  • fix(brainbar): a completed recovery clears the night; name what actually failed by @EtanHey in #1550
  • test: await owned recovery-oracle writers instead of a 3s join deadline by @EtanHey in #1557
  • fix(brainbar): nightly failure accounting follow-ups to #1550 (runner errors, recovery, deferrals) by @EtanHey in #1553
  • fix(brainbar): report one cause for an abandoned night without step details by @EtanHey in #1559
  • fix(brainbar): one status truth for icon, menu, Jobs row and Dashboard by @EtanHey in #1560
  • fix(maintenance): reclaim the weekly VACUUM WAL before releasing the maintenance hold by @EtanHey in #1554
  • fix(relay): escalate real outages only; backup alerts stay in the lead's collab by @EtanHey in #1558
  • feat(maintenance): retry failed daytime-safe steps under calendar custody by @EtanHey in #1555
  • feat(brainbar): confirm and retry failed daytime maintenance steps by @EtanHey in #1556
  • fix(brainbar): a routine partial vector scan is not a health notice by @EtanHey in #1562
  • fix(brainbar): the memory chart counts saved memories, not agent activity by @EtanHey in #1564
  • fix(updater): expect the linked brainlayer keg, not the first listed version by @EtanHey in #1561
  • fix(backup): say plainly why an unconfirmed BrainBar database blocks the backup by @EtanHey in #1563
  • fix(maintenance): independent WAL checkpoint fallback that never blocks the calendar or backup by @EtanHey in #1565
  • fix(sqlite): every writer recycles its WAL under one shared size and checkpoint policy by @EtanHey in #1566
  • chore(release): prepare BrainLayer 1.5.80 by @EtanHey in #1567

Full Changelog: v1.5.79...v1.5.80