Integrations
Connect supported alert sources, notification providers, Slack ChatOps, Jira, and signed webhooks.
OpsKnight integrations have three distinct directions:
| Direction | Purpose | Examples |
|---|---|---|
| Inbound | Normalize a provider webhook into trigger, acknowledge, or resolve actions for one service. | CloudWatch, Datadog, Prometheus, CI/CD, uptime tools. |
| Outbound notification | Deliver incident and escalation messages to responders or external endpoints. | Email, SMS, push, WhatsApp, Slack, service webhooks. |
| Workflow | Connect incident response to another working surface. | Slack ChatOps war rooms and Jira issues/action items. |
There is no native voice/PSTN notification channel in v1.4. PagerDuty support is inbound Events API v2 compatibility, not a full PagerDuty product or bidirectional synchronization.
Start here
- Read How integrations work for keys, URLs, signing, deduplication, recovery, and safe testing.
- Read the Inbound webhook reference for the exact shared authentication, signature, lifecycle, rate-limit, and response contract.
- Create the integration under Service → Integrations.
- Follow the provider runbook below.
- Trigger and resolve a synthetic alert using the exact production path.
- Configure and test outbound notification providers.
Inbound integration catalog
These entries are backed by v1.4 route handlers. The provider guide is authoritative for payload and recovery behavior.
| Category | Provider | Webhook path | Guide |
|---|---|---|---|
| APM and observability | AppDynamics | /api/integrations/appdynamics |
AppDynamics |
| APM and observability | Datadog | /api/integrations/datadog |
Datadog |
| APM and observability | Dynatrace | /api/integrations/dynatrace |
Dynatrace |
| APM and observability | Grafana | /api/integrations/grafana |
Grafana |
| APM and observability | Honeycomb | /api/integrations/honeycomb |
Honeycomb |
| APM and observability | New Relic | /api/integrations/newrelic |
New Relic |
| APM and observability | Sentry | /api/integrations/sentry |
Sentry |
| APM and observability | Splunk Observability | /api/integrations/splunk-observability |
Splunk Observability |
| Cloud | AWS CloudWatch | /api/integrations/cloudwatch |
AWS CloudWatch |
| Cloud | Azure Monitor | /api/integrations/azure |
Azure Monitor |
| Cloud | Google Cloud Monitoring | /api/integrations/google-cloud-monitoring |
Google Cloud Monitoring |
| Metrics and alerting | Icinga 2 | /api/integrations/icinga |
Icinga |
| Metrics and alerting | Nagios | /api/integrations/nagios |
Nagios |
| Metrics and alerting | Prometheus Alertmanager | /api/integrations/prometheus |
Prometheus |
| Metrics and alerting | Zabbix | /api/integrations/zabbix |
Zabbix |
| CI/CD | Bitbucket | /api/integrations/bitbucket |
Bitbucket |
| CI/CD | GitHub | /api/integrations/github |
GitHub |
| CI/CD | GitLab | /api/integrations/gitlab |
GitLab |
| CI/CD | Vercel | /api/integrations/vercel |
Vercel |
| Uptime | Better Uptime | /api/integrations/better-uptime |
Better Uptime |
| Uptime | Pingdom | /api/integrations/pingdom |
Pingdom |
| Uptime | Uptime Kuma | /api/integrations/uptime-kuma |
Uptime Kuma |
| Uptime | UptimeRobot | /api/integrations/uptimerobot |
UptimeRobot |
| Logs/events | Elastic/Kibana | /api/integrations/elastic |
Elastic/Kibana |
| Logs/events | Splunk On-Call | /api/integrations/splunk-oncall |
Splunk On-Call |
| Compatibility | PagerDuty Events API v2 | /api/integrations/pagerduty and /api/integrations/pagerduty/v2/enqueue |
PagerDuty-compatible ingest |
| Custom | Generic webhook | /api/integrations/webhook |
Custom webhooks |
The application also has /api/integrations/health for authenticated integration-health inspection. It is an operational endpoint, not an inbound alert source.
Outbound notification and workflow guides
| Capability | Guide | Scope |
|---|---|---|
| Email, SMS, push, WhatsApp, provider tests, and history | Notifications | Workspace providers and user preferences. |
| Slack notifications | Slack | Workspace connection and incident messages. |
| Slack OAuth | Slack OAuth setup | App credentials, scopes, redirect, and connection. |
| Slack ChatOps and war rooms | Slack ChatOps | Dedicated incident channels, commands, and actions. |
| Jira Cloud | Jira | Workspace configuration, service mapping, incident/action-item links. |
| Service webhooks | Custom webhooks | Outbound lifecycle webhook configuration and signing. |
Microsoft Teams and Google Chat do not have dedicated native notification providers in the v1.4 provider model. A compatible incoming-webhook endpoint may accept a generic outbound webhook payload, but test its format explicitly and do not describe it as a native integration.
Credential vocabulary
| Credential | Used for | Created where |
|---|---|---|
| Integration ID | Selects the service integration record on provider-native webhook routes. | Service integration record. |
| Integration key / routing key | Authorizes inbound provider-native and Events API traffic and routes it to a service. | Generated when the integration is created. |
| Signature secret | Optionally verifies the raw provider webhook with the route's supported HMAC scheme. | Rotated/cleared on the integration. |
| Workspace API key | Authorizes deliberately published API operations according to scopes. | Settings → API Keys. |
| Outbound webhook secret | Signs OpsKnight-to-consumer requests. | Generated with the webhook configuration. |
Never substitute one credential type for another. Keep them in secret storage and rotate after exposure.
Production acceptance
An integration is ready only when:
- the provider can authenticate to the exact route;
- a representative failure creates the expected service incident and urgency;
- a repeated failure reuses the same deduplication key;
- recovery resolves that incident when supported;
- invalid key and invalid signature requests are rejected;
- a disabled integration rejects or stops processing as documented;
- rate limiting and retry behavior are understood;
- responders receive the intended outbound notifications;
- credential rotation and removal have been rehearsed.
Troubleshooting
Start with Event Logs, notification history, the incident timeline, and Troubleshooting. A successful provider delivery only proves OpsKnight accepted the request; it does not prove escalation or outbound provider delivery succeeded.
Last updated for v1.4
Edit this page on GitHub