In `workbuddy2api-panel`, setting `admin.audit_enabled: true` in `config.json` while leaving `admin.enabled: false` causes a fatal config validation error on st…
In workbuddy2api-panel, setting admin.audit_enabled: true in config.json while leaving admin.enabled: false causes a fatal config validation error on startup and puts the container into a Restarting (1) crash loop.
admin.enabled only controls registering POST /admin/tasks/{name}/run, and admin.audit_enabled attaches the JSONL audit middleware strictly to that /admin/ route. Config loading rejects admin.audit_enabled=true when admin.enabled=false, while /panel/ itself does not use admin.* (it reuses api_key).
Set both admin.enabled and admin.audit_enabled to false unless /admin/tasks/{name}/run is explicitly required, in which case enable both together. Verify config before recreating the container.
Container starts cleanly, passes /healthz, and does not emit load config: admin.audit_enabled=true in docker logs.
docker ps shows container healthy/running (not Restarting) and curl http://127.0.0.1:7863/healthz returns 200.