Examples

Things to ask once your assistant is connected:

Triage

• "How many critical alerts did we get in the last 24 hours, and on which hosts?"
• "Give me a severity breakdown of this week's alerts for the production workspace."
• "Are there any active container_escape or vm_escape alerts right now?"

Investigation

• "Show me the details and timeline of the most recent gpu_ransomware alert."
• "What alerts has gpu-node-17 generated today? Walk me through the process timeline of the worst one."

Host activity

• "What else happened on gpu-node-17 around its latest critical alert?"
• "Which programs ran only once around the most recent container_escape alert?"
• "Who started the process flagged in the most recent gpu_ransomware alert, and what else did its parent run?"
• "Did anything else on gpu-node-17 open /etc/shadow around its latest critical alert?"

GPU health

• "Which GPUs aren't healthy right now, and what's their temperature?"
• "Give me the full picture on GPU-8f2c… — specs, current load, and any security findings."
• "How is alert volume trending over the past week? Anything spiking?"

Reporting

• "Which workspace is generating the most alerts this month, and of what types?"
• "Summarize this week's security posture across all our GPUs in a few sentences."