Using Agents & Tools
The Agents page shows all AI agents working on your behalf - their health, status, and activity.
Agent Types
System Agents
RubixKube Observer
Role: Monitors your cluster continuouslyWhat it does:
- Watches pod health
- Collects metrics
- Detects failures
- Sends events to RCA Pipeline
RCA Pipeline Agent
Role: Analyzes incidents and generates RCA reportsWhat it does:
- Receives events from Observer
- Gathers logs and metrics
- Correlates patterns
- Generates root cause reports
Memory Agent
Role: Stores and learns from incidentsWhat it does:
- Stores RCA reports
- Builds knowledge graph
- Recalls similar past incidents
- Improves over time
SRI Agent
Role: Powers the Chat interfaceWhat it does:
- Answers your questions
- Queries cluster data
- Provides recommendations
- Explains incidents
Agent Health Status
| Status | Meaning | What to Do |
|---|---|---|
| ✅ Active | Agent running normally | Nothing - all good |
| ⚠️ Degraded | Slow or intermittent | Monitor, may self-recover |
| ❌ Offline | Agent not responding | Check pod, restart if needed |
Monitoring Agents
Total Agents: Should match your deployment (usually 3-4) System Agents: Core RubixKube agents Active Agents: Currently operational If an agent goes offline:- Check the Agents page
- Identify which one
- Click “View” for logs
- Restart pod if necessary
What You Learned
✅ 4 core agents✅ Their roles
✅ Health statuses
✅ What to do when offline