Alert-based Debugging enables AI to begin an investigation using the alert context as the starting point. When you click “Debug with AI” on an alert in the Alerts Inbox, the platform opens the Alert Details page where the agent immediately initiates a structured investigation.
This mode does not require the user to input anything — the AI uses available metadata, tags, and context from the alert to propose likely causes and remediation paths.
Each root cause goes through a lifecycle of verification:
Status | Description |
---|---|
Pending | The hypothesis is created but not yet verified (often due to no integrations) |
Evaluating | The agent is currently verifying the root cause by analyzing relevant data |
Unverified | Agent attempted verification but couldn’t complete (e.g., permission errors, timeout) |
Ruled Out | The agent checked and confirmed that this cause is not valid |
Detected | The hypothesis was verified and is considered the most likely root cause |
Alert-based Debugging enables AI to begin an investigation using the alert context as the starting point. When you click “Debug with AI” on an alert in the Alerts Inbox, the platform opens the Alert Details page where the agent immediately initiates a structured investigation.
This mode does not require the user to input anything — the AI uses available metadata, tags, and context from the alert to propose likely causes and remediation paths.
Each root cause goes through a lifecycle of verification:
Status | Description |
---|---|
Pending | The hypothesis is created but not yet verified (often due to no integrations) |
Evaluating | The agent is currently verifying the root cause by analyzing relevant data |
Unverified | Agent attempted verification but couldn’t complete (e.g., permission errors, timeout) |
Ruled Out | The agent checked and confirmed that this cause is not valid |
Detected | The hypothesis was verified and is considered the most likely root cause |