[Feature]: Surface trigger-level failures — alert on connection ERROR, show failed trigger runs in Issues
Author: majewskibartoszCreated Jul 20, 2026Updated Sep 16, 2026
Labels🌟 feature🔄 area/flows💫 priority
Problem
OAuth refresh fails → connection flips to ERROR (app-connection.handler.ts:184, status update only, no side effects). From then on, every incoming webhook fails silently:
- Trigger hook throws ConnectionExpired → no flow run created; payloads only submitted on OK (
execute-webhook.ts:134) - No issue, no alert — alerts fire only from flow-run hooks (
flow-run-hooks.ts:38); no flow run = nothing fires - Only signals: connection badge on Connections page + platform-admin trigger-health aggregates (per-piece counts, not per-flow)
- Payloads lost once queue retries exhaust
Verified on current main.
Ref: Pylon 2516
Proposed
Needs product scoping (Mo: per-flow Trigger Run log in UI, hidden by default; treating trigger failures as flow runs breaks retry — one payload maps to many runs, formats differ). Minimum viable:
- Alert (email/notification) when a connection transitions to ERROR
- Per-flow visibility of failed trigger runs — Issues entry or Trigger Runs list with a "connection error — reconnect" CTA
Prior art
- Customer PR (closed, not merged): https://github.com/activepieces/activepieces/pull/8726 — returned raw payload on trigger failure so a failed run gets created
- A
TriggerRunentity existed and was removed (migrationRemoveTriggerRunEntity, Oct 2025);recordTriggerRunnow keeps only per-piece platform aggregatesactivepieces/activepieces#13994
Source: activepieces/activepieces