test urgent with agent

OPS-39 agent-ops Canceled urgent

Adopt successor:"@next" on gitpush-core.sh's push-gate unlock, or keep the bare unlock - the FIFO queue is non-binding without it

OPS-286 agent-ops Canceled high

llmmsg-srv roster & PM resilience: pruneStale soft-offline, offline buffering, PM re-election

OPS-54 agent-ops Canceled high

backup-whey.service failing since 2026-05-09 02:03

OPS-31 agent-ops Canceled high

one-db: consolidate all whey telemetry/KPI sqlite into one local Postgres 17 db (telemetry), schema-per-domain, SELECT*-friendly views, reporting web UI. Goal: one place for all KPIs/perf/context data + Elazar can actually find/read them.

OPS-2 agent-ops Canceled high

└─ one-db: views layer — every table gets enough pg views (latest / daily / report rollups) so agents query a view with SELECT * (never name fields). Goal: zero token waste guessing columns. Document view list in telemetry-inventory.md.

OPS-4 agent-ops Canceled high

└─ one-db: cutover (NO dual-write — Elazar dropped it 2026-06-06). Collectors write straight to pg via metrics_write (spools to JSONL on pg-down, so no data loss); backfill already loaded full history. Flip 5 collectors + netreport off sqlite, retire sqlite files once pg-proven.

OPS-6 agent-ops Canceled high

└─ one-db: lazy-load telemetry-inventory.md — single doc agents read on demand listing, per data stream: which systemd unit/timer/cron collects it, which script the unit runs, which table it writes, and which script reports to Elazar + which aro. Keep it the SSOT map of the whole telemetry pipeline.

OPS-3 agent-ops Canceled high

└─ one-db: kpi.pensanta.com web UI on whey — small read-only server (chat-duo-web model, single-file .mjs) that reads the SELECT* views and renders KPI pages, so Elazar has one habit-friendly place to glance at all KPIs and drill into any metric when something looks off.

OPS-5 agent-ops Canceled

Alert windows that straddle a host reboot are reported as agent faults: no-statusline duration spans downtime the process did not exist for

OPS-126 agent-ops Canceled

context-monitor.sh reports stale ctx for idle whey agents: frozen transcript pins them at >=180k for hours (OPS-11 residual, action gated but read still wrong)

OPS-124 agent-ops Canceled

Commanded /compact does not land: 3 agents measured unverified in one pass, root cause unknown

OPS-107 agent-ops Canceled

Stuck-busy alert has no known-wedged latch: re-fires every 6h forever with no escalate-once path

OPS-122 agent-ops Canceled

Right-size event_log telemetry retention: 7d raw is ~800MB steady state, gates_evaluated JSON is a third of it

OPS-120 agent-ops Canceled

cc-context efficiency report - read-only reporting layer on monitor-context.sqlite (week/day/48h)

OPS-60 agent-ops Canceled

context-monitor.sh 180k nudge DM promises a 220k injection that cannot happen under COMPACT_ACTUATOR=0

OPS-110 agent-ops Canceled

scan-context coverage alarm is permanent noise: Codex audit-*-ca panes have no CC statusline to read

OPS-109 agent-ops Canceled

cc-context-monitor zombie-detector false-positives on reactive-idle agents: rule (turns>=2 && tool_calls==0 -> zombie) can't distinguish a self-armed-wakeup zombie from a reactive agent woken by inbound llmmsg DM that acked + stood down with no tools needed. At 05:00 it flagged audit-venus-cc (benign: empty CronList, zero ScheduleWakeup ever, 100% inbound-DM-driven) -> 2 dup alerts + cross-host investigation. Hits EVERY reactive agent (host nw between alerts, audit agents). Fix (nw-whey suggested): a wake is zombie-suspect only if NOT preceded by an inbound message in the window (self-triggered) AND/OR non-empty CronList; if wake correlates with inbound DM + empty CronList -> reactive-idle, suppress. Hub has inbound-msg timestamps to correlate. Related: OPS-11 last_turn stale-read.

OPS-17 agent-ops Canceled

whey's sys-check drop-in skips pg on whey because 'whey has ZERO telemetry surface' - whey now writes telemetry to venus pg and the premise has expired silently

OPS-218 agent-ops Canceled

sys-check.sh's pg skip-list defaults to 'lezama venus' and venus pg is answering - the default is measurably stale

OPS-214 agent-ops Canceled

STRUCTURAL: /loop self-/compact over-cap workstream. Proven (Elazar 2026-05-21, ~28-30k ctx): a /loop session self-/compacts via ScheduleWakeup prompt /compact. Replaces cap->park->relaunch with cap->ss->self-/compact->continue. Gated on a ccs.sh launcher decision (ScheduleWakeup exists only in /loop-mode sessions) — needs Elazar GO. Separate from v2.7 over-cap path

OPS-51 agent-ops Canceled

Scope a HARDWARE remote-reset path for dead-kernel recovery (autossh/userland watchdog cannot revive a dead kernel): IPMI/iDRAC if present, else smart PDU - only no-physical-access option

OPS-84 agent-ops Canceled ↳ OPS-80

ZeroTier transport redundancy: zt_whey_visible=0 across all 5435 probes -> single public-dns path. Bring ZT up (zerotier-cli listpeers, whey visible) as a real fallback route

OPS-83 agent-ops Canceled ↳ OPS-80

Per-ccs maintainer & script/systemd ownership registry

OPS-26 agent-ops Canceled

└─ Authority matrix: per script/unit, who maintains, monitors, modifies, notifies. Decide: DB-backed (bwi) vs file-backed (~/.claude/maintainer-agents.md) source of truth.

OPS-28 agent-ops Canceled

└─ Process for new scripts: who creates, who maintains, alert routing. Capture in maintainer-agents.md once authority matrix decided.

OPS-30 agent-ops Canceled low

Decide and execute: context-monitor.sh --drain path + its timer are undocumented, possibly dead

OPS-119 agent-ops Canceled low

Retire-vs-cutover: context-venus-monitor-alert.sh is a dormant duplicate of the live context-monitor-alert.sh

OPS-118 agent-ops Canceled low