Integrations

GridRouter alerts you where your team already works, and sends signed webhooks to your own systems when calls, jobs, runs, budgets and access change.

Alert destinations

Route each alert rule to every destination or to the ones you pick, with a minimum severity per destination.

Subscribe an HTTPS endpoint to the events you need, or to all of them.

Calls
- call.completedA call finished with a hit or a miss.
- call.failedA call failed.
Jobs and lists
- job.completedAn async job succeeded.
- job.failedAn async job failed or expired.
- list.completedA list finished processing every row.
Waterfalls
- waterfall.run.completedA waterfall run finished.
Alerts
- alert.triggeredAn alert rule fired.
- alert.resolvedAn alert resolved.
Vendors and billing
- vendor_key.failedA vendor rejected one of your own vendor keys.
- billing.mismatch_detectedA vendor billed a call against its own published rule.
- budget.threshold_reachedA budget crossed a threshold.
Access
- key.createdAn API key was created.
- key.revokedAn API key was revoked.
- member.addedSomeone joined the workspace.
- member.removedSomeone left or was removed from the workspace.

Signed
Every webhook carries Standard Webhooks headers: an id, a timestamp and an HMAC-SHA256 signature. Rotate the secret with a day of overlap.
Retried
Up to 8 attempts over about 23.6 hours, with the same webhook id every time so you can dedupe. An endpoint that fails for days is disabled and you get an alert.
Logged and replayable
Every attempt's status, latency and response snippet, kept 30 days. Replay any delivery or send a test event. Events over 256 KB arrive as a stub with the ids to fetch.
Secrets stay write-only
Webhook URLs, routing keys and API keys are encrypted in your workspace's vault and never shown again. Outbound URLs must be public HTTPS and redirects are never followed.

Every new workspace starts with alerts for failing vendor keys, budgets, billing mismatches, error spikes and failing deliveries.