First Steps Guide
You have Alga running. Here's what to do next.
1. Complete First-Run Setup
Open http://localhost:3000 in your browser. When no users exist in the database, Alga automatically redirects to the setup wizard. Enter an email, password, and full name to create the initial admin account. This is the only time the setup wizard is available — once an admin exists, it is disabled.
2. Create a Webhook Token
Go to Settings → Webhook Tokens and create a token. You'll need this to send alerts.
3. Send a Test Alert
curl -X POST http://localhost:8080/webhooks/alerts \
-H "Authorization: Bearer alga_YOUR_TOKEN" \
-H "Content-Type: application/json" \
-d '{
"alerts": [{
"status": "firing",
"labels": {
"alertname": "TestAlert",
"severity": "warning",
"namespace": "default"
},
"annotations": {
"summary": "This is a test alert"
},
"generatorURL": "http://localhost:3000/test"
}]
}'You should see the alert appear on the Alerts page.
4. Connect Grafana
- In Alga, copy your webhook token
- In Grafana, go to Alerting → Contact points → Add contact point
- Set type to Webhook
- Set URL to
http://your-alga-host:8080/webhooks/alerts?token=alga_YOUR_TOKEN - Save and test
5. Set Up Routing Rules
Go to Routes and create rules to route alerts to specific channels:
- Click Add Rule
- Set conditions (e.g.,
namespace = production) - Set destination channels (Slack or Mattermost)
- Save
Alerts that don't match any rule go to the default channel.
6. Create Knowledge Notes
Knowledge notes capture operational runbooks and context so your team (and AI agents) know exactly what to do when an alert fires.
- Go to Knowledge in the sidebar
- Create notes with tags and selectors matching your alert labels (e.g.,
alertname: HighMemory) - AI agents and operators can reference these during investigations
Knowledge notes help standardize incident response and reduce mean time to resolution. For structured, step-by-step procedures, see Playbooks.
7. Explore Core Features
| Feature | Where | What |
|---|---|---|
| Alerts | Alerts page | View, acknowledge, resolve alerts |
| Investigations | Investigations page | AI-powered root cause analysis |
| Knowledge | Knowledge page | Shared notes for operators and agents |
| Routing | Routes page | Alert routing rules |
8. Set Up Incident Management
- Create Services in the Service Catalog
- Define Dependencies between services
- Create Teams and add members
- Set up On-Call Schedules with rotations
- Create Escalation Policies with tiers and delays
9. Set Up On-Call and Escalation
On-call schedules ensure the right person is always reachable. Escalation policies define what happens when they don't respond.
- Go to On-Call → Schedules and create a rotation
- Add layers for follow-the-sun coverage if you have multiple time zones
- Create an Escalation Policy with tiered levels and delays
- Assign the escalation policy to a service (target a team or user within each level)
See Teams & On-Call for detailed schedule and escalation configuration.
10. Connect an Agent (Optional)
To enable automated investigations:
- Go to Agents in the sidebar menu and create an agent token
- Install the Hermes adapter or OpenClaw plugin
- Configure the agent to connect to
http://your-alga-host:8080/api/v1/agent/eventswith the agent token
See AI Investigation for details.
11. Configure Notifications
Each user can set their own notification preferences for how they receive alerts:
- Click your profile avatar → Notification Preferences
- Add rules mapping notification types to channels (in-app, email, Slack DM, voice)
- Set a default channel for notification types without an explicit rule
- Optionally enable the voice opt-out to suppress phone calls
- Link your personal Slack account for DM delivery under Connected Accounts
See Notification Preferences for details on all available channels.
12. Secure Your Instance
For production deployments:
- Set
ENVIRONMENT=production - Configure
ENCRYPTION_KEYSandSECRET_PEPPER - Enable
SECURE_COOKIES=truewith HTTPS - Complete the first-run setup wizard to create the initial admin account before enabling production
- See Security & Authentication for details