Default alert rules, cron schedules, workflow toggles, and health checks for self-hosted FluxyChat operators.

Observability & scheduled jobs

Default alert rules

New projects receive seeded operational alert rules via insertNewProject → seedDefaultAlertRules.

Console: Analytics → alerts list; open alerts also surface in the red System status banner (owner/admin JWT).

Mode	`WORKFLOW_SCHEDULES_ENABLED`	What runs
Production default (repo)	`false`	Worker `[triggers] crons` in `wrangler.toml`
Workflows (optional)	`true`	`FluxyScheduledWorkflow` bindings; Worker `scheduled` is a no-op

Crons when enabled (either path):

Schedule	Job
`0 8 * * *`	Daily AI digest
`/15 * * *`	Quiet-hours notification batch
`0 3 * * *`	Data retention purge
`/5 * * *`	Pending webhook delivery flush

In apps/worker/.dev.vars:

WORKFLOW_SCHEDULES_ENABLED=false

Uses Worker cron from wrangler.toml [triggers] during wrangler dev.

Manual trigger (owner/admin JWT):

POST /admin/scheduled/run
Content-Type: application/json

{ "cron": "0 8 * * *" }

Allowed values: 0 8 * * *, */15 * * * *, 0 3 * * *, */5 * * * *.

See Scheduled workflows for the Workflows setup.

Public: dashboard /status → Worker /health
Operator: pnpm --filter @fluxy-chat/worker smoke:remote -- --base-url <worker-url> --admin-jwt <token>