Skip to content
English
Do you need help? Customer portal
neowit.io
  • There are no suggestions because the search field is empty.

Integration disconnected

What the status badges mean on Settings → Integrations, how to get an integration back to Connected, and what happens to workflow executions while it's down. For admins when Slack or Philips Hue stops working, and for workflow owners trying to diagnose a red run.

What the badge means

On each integration card:

  • Connected (teal) — the last health check passed. Actions work.
  • Bad credentials (red) — the integration's token was rejected on the last check. Usually because the admin on the other side (Slack workspace admin, Hue account holder) revoked Neowit's access, or the token expired beyond auto-refresh.
  • Unable to connect (red) — Neowit couldn't reach the provider's API on the last check. Usually a transient network or cloud-outage issue.
  • New (grey) — a fresh integration that hasn't reached Connected yet. Finish setup and wait for the first health check.
  • Deactivated (red) — rare. If you see it, contact support.

[SCREENSHOT] The Integrations settings page with three rows: one Slack with a teal Connected badge, one Slack with a red Bad credentials badge, and one Philips Hue with a red Unable to connect badge.

Why the badge can lag reality

Status is refreshed by a health check that runs every 30 minutes. If a token is revoked at 10:00, the badge may not flip to Bad credentials until 10:30 — but actions already fail during that window. The badge is just slow to update.

On the Hue integration, click Synchronize to force a check now. Slack has no such button; you'll wait for the next cycle.

Recover from "Bad credentials"

Slack

  1. Open Settings → Integrations → Slack.
  2. If you just need to freshen the token, paste a new API Token into the edit form and click Save.
  3. If you want the OAuth flow again (simpler for most people): delete the integration from the Integrations list, then click Add integration → Slack and re-run the OAuth. Existing workflows keep working after the new integration is in place — they reference node types, not a specific integration record.

Philips Hue

  1. Open Settings → Integrations → Philips Hue.
  2. Click Connect to Philips Hue on the edit page to re-run OAuth.
  3. Sign back in with the Hue account that owns the bridge and approve Neowit's access.

Recover from "Unable to connect"

Usually transient. Try in order:

  1. Wait a few minutes — the next health check may clear it on its own.
  2. For Hue, click Synchronize to force a check now.
  3. Check the provider's status page (Slack and Philips both publish one).
  4. If it persists for an hour or more with no outage reported, treat it as Bad credentials and reconnect.

What happens to workflows while an integration is down

  • Triggers still fire. Workflow executions start as normal — the integration's state doesn't block trigger routing.
  • Actions fail fast. The first task that calls the disconnected integration returns "Not authenticated" and the workflow stops there.
  • There is no queue or retry. Once an action has failed, it doesn't run again automatically when the integration recovers. The failed execution stays red in the history.
  • Downstream actions don't run. A failed task halts the workflow.

If you have executions you wanted to happen during the outage, you'll need to re-trigger them manually — re-emit the source event, or re-run the workflow from the editor via Run workflow with a synthetic payload.

 

Tip: If an integration is going to be down for a while, flip the workflow's Enabled toggle off in the Save dialog. That keeps the execution history clean and saves you a stream of red runs to ignore later.

Delete and re-add

When editing the integration doesn't recover it — or when the provider has fully revoked access and you need a fresh OAuth grant — delete and re-add:

  1. Go to Settings → Integrations.
  2. Select the integration's row and click Delete in the toolbar.
  3. Click Add integration and re-run the setup flow.

Workflows that use that integration's actions don't need to be rebuilt — they reference node types, not specific integration records. Once the new integration is in place, the next trigger runs against it.

Integration vs. workflow — two different toggles

If runs aren't happening at all, figure out which switch is off:

  • An integration in red state means actions fail, but triggers still route and executions show in the history (marked red).
  • A workflow with Enabled off means triggers don't route at all, and nothing appears in the history.

See Workflow didn't run for the case where even the red executions aren't appearing.

Related