← back to Chief Of Operations

README.md

24 lines

# Chief of Operations

Local read-first harness for keeping the terminal operation on track.

It checks the master `dot-screen-router`, the `desktop-dotbar` launch agent and health endpoint, Plannator (`:9772`), semantic terminal status, router evidence (`green_misplaced` / `attn_misplaced`), screenshot freshness, and archived legacy arrangers.

```sh
./chief_ops.py check
./chief_ops.py arrange
./chief_ops.py serve # http://127.0.0.1:9896
```

The Arrange action delegates to the existing master router through the desktop bar API. The harness does not delete, publish, or mutate fleet data.

## Continuous ticket supervision

`supervisor_loop.sh` refreshes `state/queue.json` every 60 seconds under launchd (`com.steve.chief-of-operations`). Each pass also runs the Exo watchdog, which probes all configured hosts three times, writes `state/exo-watch.json`, and logs a ticket event when the ring changes to partial or offline. The queue includes open, blocked, and doing tickets, scores Designer Wallcoverings/client-impact work first, and records the next safe step. It automatically claims one highest-priority open ticket at a time as `chief-of-operations`; blocked tickets remain surfaced for blocker review.

```sh
./chief_ops.py queue
./chief_ops.py claim
./chief_ops.py watch-exo
```