Change feed
What changed, in one place — the first question of every incident.
Settings → Platform → Change Feed is one chronological list of everything that changed.
Visible to everyone
Deliberately not admin-only.
"What changed?" is the first question during an incident. Gating it behind
admin would make the on-call engineer wake somebody else to ask it.
What is in it
| Change | From |
|---|---|
| Deployments | Delivery |
| Configuration changes | Settings, thresholds, routing |
| Feature-flag flips | Feature flags |
| Cloudflare traffic shifts | Load balancing |
| Maintenance windows | Maintenance |
| Rule changes | Created, edited, disabled |
| Datasource changes | Credentials, endpoints |
Using it
Snap the range to the incident window. What is in the list is what to look at first.
The feed says what changed. It does not say what caused the problem, and it never claims to. During a busy hour several unrelated things change, and the honest statement is the temporal one — the judgement is yours.
Config inspector
Settings → Platform → Config Inspector answers a narrower question: what was this setting at 14:00?
Configuration is snapshotted, so you can compare a setting now against what it was before the incident. "It has always been like that" is frequently wrong and this is how you find out.
Related
- Unified timeline — the same idea with alerts and incidents on the same axis
- Audit log — who did it, rather than what happened
- Copilot
correlate_window— all of it, from one question
Where this behaviour lives: backend/internal/worker/change_collector_worker.go, frontend/src/app/changes/. If the code and this page disagree, the code is right — please fix the page.
Part of Administration — Settings, retention, platform thresholds and change tracking.