Integrations

The two vendors brs replays directly, and the wider set the corpus tags a signature against.

A replay comes from one vendor at a time, and today brs replay executes two: Prometheus (the cross-vendor #1) and Datadog composite monitors (cross-vendor #2, PLAN-FASE-06). Both produce the same classified report shape, so a rule migrated from one to the other can be replayed against its own history either way.

Prometheus

Needs a TSDB snapshot, a rules.yml, and optionally an alertmanager.yml — see CLI reference for every flag.

Terminal

brs replay --tsdb ./snapshot --rules rules.yml --alertmanager alertmanager.yml

Datadog

Needs one export bundle — monitors and metric series together — and cannot be combined with the Prometheus flags:

Terminal

brs replay --datadog ./export.json

Vendor scope in the corpus

A corpus signature's vendorScope can name a broader set — prometheus, grafana, grafana_cloud, mimir, victoriametrics, datadog, newrelic — because a failure signature is portable across any Prometheus-remote-write-compatible backend even when the replay engine only executes two of them directly today. Search the public corpus (/corpus) by vendor to see which classes are documented for a backend brs does not replay yet.

PR-gate

Either vendor's rules can be gated on every pull request without running brs by hand — see PR gate.