Integrations

Provider directory

• Slack routes incidents and escalation updates into response channels.
• PagerDuty and Opsgenie connect incident notifications to on-call systems.
• Jira creates and links follow-up work from incidents and reviews.
• GitHub supports deployment and change correlation workflows.
• Datadog connects monitoring events to existing telemetry dashboards.
• Webhook sends normalized events to customer-owned automation endpoints.

Governance

• Only users with integration management permission can connect or disable providers.
• Integration connection and disable actions are audit logged.
• Provider state is organization-scoped.
• Use least-privilege API tokens and provider-specific secrets when wiring production connections.

Implementation model

• The product stores provider, display name, status, configuration metadata, creator information, and delivery evidence.
• Provider status is explicit: not connected, configured, connected, failed, or disabled.
• Configured means setup is saved but not verified yet.
• Connected means the provider accepted a delivery or, for GitHub, sent a valid signed webhook.
• Failed means delivery exhausted retries or the provider rejected the request.
• Connection status can be disabled without deleting operational history.
• Delivery logs show event type, attempt count, provider status, message, and timestamp.
• Use webhooks for providers that are not yet first-class integrations.

GitHub setup

• Paste the repository as owner/repository or use the full GitHub repository URL.
• Use the AImonitoring GitHub webhook URL shown on the Integrations page.
• Set GitHub webhook content type to application/json.
• Use the same webhook secret in GitHub and AImonitoring so signatures can be verified.
• Subscribe to push, release, deployment, deployment_status, and workflow_run events.
• GitHub token is optional for webhook correlation; add a token only when issue creation or API sync is needed.

Related documentation