Watchdog: verify standup cron recovery (May 8-11 outage)

Pre-committed watchdog from LEARNINGS.md 2026-05-11 entry. Background: Daily Standup cron failed May 8 with `cron payload.model anthropic/claude-sonnet-4-5 rejected by agents.defaults.models allowlist`. Self-Reflection cron had the same misconfig and skipped May 8/9/10 nights. My May 8 heartbeat fix attempt edited ~/.openclaw/cron/jobs.json but did not verify the next cycle; problem persisted 3 more days. Tonight (May 11 04:40 UTC) self-reflection finally fired because the model field is now `anthropic/claude-opus-4-7` (different value from what I wrote May 8). Verification conditions for this watchdog: 1. Check if /home/agent/agents/artem-agent/memory/standups/2026-05-11.md exists by 2026-05-11 12:00 UTC. If yes → standup cron is recovered, mark this task completed with note `standup cron recovered May 11`. 2. If /home/agent/agents/artem-agent/memory/standups/2026-05-11.md is MISSING by 2026-05-11 12:00 UTC, execute the active-push protocol: a. PATCH this task to status=blocked with reason `standup cron still broken May 11 (4 consecutive days)`. b. Send Slack DM to Artem: `Standup cron has been down since May 8 (model allowlist misconfig in ~/.openclaw/cron/jobs.json). I tried to fix it on May 8 but the fix did not propagate. May 11 standup also missing. Could you check the cron config or restart whatever reloads it? Self-reflection cron recovered tonight on its own. Full timeline in memory/2026-05-11.md.` c. File a follow-up fleet-task to track resolution. Reference: LEARNINGS.md 2026-05-11 entry titled `Verify Infrastructure Fixes The Next Cycle; A Pre-Committed Re-Open Trigger Is Worthless Without A Session To Execute It`. memory/2026-05-11.md has the full timeline.