Overview
Mission Control
One clean surface for team status, recent work, pending owner decisions, and rollout governance.
Agents
13
1 command lane + 5 specialist lanes
Active lanes
8
0 persistent-ready
Pending decisions
0
Owner-facing items waiting on Zo
Tasks completed
19432
No active tasks
Agent roster
Current lane state at a glance
Derreck
Commandpersistent activeChief of Staff / Orchestrator
Forge
BuildarchivedSenior Dev Lead
Mace
engineeringpersistent activeSenior Dev Lead
Ledger
Mainstay OpsarchivedMainstay Ops Lead
Cash
financepersistent activeSettlement & Finance
Anchor
ReliabilityarchivedOps / Infrastructure Lead
Blitz
growthpersistent activeGrowth & Marketing
Atlas
PlanningarchivedResearch / Planning Lead
Iris
operationspersistent activeOps & QA
Signal
GrowtharchivedGrowth / Content Lead
Sage
researchpersistent activeResearch & Strategy
Pixel
creativepersistent activeCreative Writer
Donald
businesspersistent activeBusiness Manager
Pending decisions
What still needs owner attention
No pending decisions
Watch items
Operational notes worth keeping visible
- 5 review checkpoint(s) pending
Recent work
Latest tasks across all lanes
Proactive Sweep
SWEEP_OK — enabled cron trio is healthy, containers/services/disk/GPU are normal, older non-ok cron residues are confined to disabled Tilly jobs, and the only open items remain the same 2 owner-gated approvals already queued in `ops/decision_queue.json`.
Self Health Check
HEALTH_OK — active cron trio is healthy on raw JSONL evidence, all enabled jobs use allowed `openai/gpt-5.4` overrides, Blitz memory is intact, the queue is at 2 approval-gated items, expected containers are up, and `/` is at 13% used. One caveat I verified and documented: this host rejects `journalctl --since "today 00:00"` with `Failed to parse timestamp: today 00:00`, so I reran the auth check with `--since "2026-06-28 00:00:00"`. That grep produced only a local websocket `unauthorized ... reason=token_missing` line from `openclaw-control-ui`, not a `401` or token-refresh error, so there is no evidence of an OpenAI auth failure. I recorded the run in [memory/2026-06-28.md](/home/xclusivesystems/.openclaw/workspace/memory/2026-06-28.md:1) and closed Control Center task `cdf282a0-c6c1-4ab7-b062-8de1fd0745e9` as `verified_done` against `http://127.0.0.1:15100/api/system`.
Proactive Sweep
SWEEP_OK — active cron set is healthy, Control Center/Red/CRM/LightRAG endpoints are up, containers are healthy, disk is 13%, GPU is normal, and the only open items remain the same 2 owner-gated approvals already parked in `ops/decision_queue.json`; I refreshed [memory/2026-06-28.md](/home/xclusivesystems/.openclaw/workspace/memory/2026-06-28.md:1) and [SESSION-STATE.md](/home/xclusivesystems/.openclaw/workspace/SESSION-STATE.md:1).
Proactive Sweep
SWEEP_OK — active cron trio has fresh `status=ok` runs, Red/CRM/Control Center/LightRAG health checks all returned `200`, expected containers are up, no failed systemd units are present, disk is at `13%`, RTX 5080 is normal at `34 C / 0%`, and I logged the 14:00 ET sweep in [memory/2026-06-28.md](/home/xclusivesystems/.openclaw/workspace/memory/2026-06-28.md:1) plus refreshed [SESSION-STATE.md](/home/xclusivesystems/.openclaw/workspace/SESSION-STATE.md:1).
Proactive Sweep
SWEEP_OK — Red, CRM, Control Center, and all 3 LightRAG instances are healthy; expected containers are up; disk is 13%; RTX 5080 is normal at 33 C/0% util; gateway auth is clean; the only recent cron anomaly was a one-off `proactive-sweep` interruption during a gateway restart and this sweep is the recovery cycle; the only pending items remain the 2 existing owner-gated approvals already parked in `ops/decision_queue.json`.