Option A: Connect your existing tools (recommended)
If you already use Sentry, Datadog, New Relic, or another observability tool, Rocketgraph reads from them directly. No SDK changes needed.1. Sign up and connect a data source
Go to rocketgraph.app/settings and connect your observability tool:- Sentry — Org slug, project slug(s), and a User Auth Token
- Datadog — API key and Application key
- New Relic — API key and Account ID
- CloudWatch — AWS access key, secret, region, and log group
2. Set up deploy webhooks
Connect your deploy pipeline so Rocketgraph runs a before/after analysis on every deploy:- Railway — Copy the webhook URL from Settings and add it in Railway project → Settings → Deploy Webhooks
- GitHub Actions — POST to the webhook URL on push to main
3. Connect Slack
Go to rocketgraph.app/integrations/slack and install the Rocketgraph bot. Anomaly reports and deploy regression alerts are posted to your channel.4. Deploy something
Push a change to your main branch. After the deploy, Rocketgraph will:- Fetch logs from before and after the deploy
- Run ML-based template mining and anomaly scoring
- Post a diff to Slack: new errors, disappeared patterns, behavior changes
Option B: Send OpenTelemetry directly
If you want to send traces, metrics, and logs directly to Rocketgraph (instead of or alongside your existing tools):AI agent (recommended)
The fastest way to instrument a Node.js project:Manual setup
Set environment variables and use the OpenTelemetry auto-instrumentation:Next steps
Anomaly detection
How continuous and deploy-time anomaly detection works.
Sentry integration
Connect Sentry for error tracking across all your services.
Railway / GitHub
Deploy webhooks for regression detection.
AI SRE in Slack
Get anomaly alerts with root cause analysis.