Skip to main content

Autodiagnostics

Autodiagnostics generates an AI-powered diagnosis for any resource managed by SleakOps. The AI reads logs and infrastructure state, identifies the root cause, and suggests corrective actions — so you spend less time debugging and more time shipping.

autodiagnostics-demo

FAQs​

What is Autodiagnostics?​

Autodiagnostics is an AI agent embedded in SleakOps that analyzes any resource — a failing deployment, a crashing Pod, a dependency with connectivity issues — and produces a structured diagnosis: what went wrong, why it happened, and what to do about it.

Instead of manually correlating logs, Kubernetes events, and infrastructure state, you get a clear explanation directly inside the console.

Where do I trigger a diagnosis?​

The Diagnose action appears on individual resource detail pages when there is an issue to analyze:

  • Deployments and releases
  • Projects and services
  • Dependencies

Once triggered, the diagnosis opens as a conversation in the Support section (/support/) where you can review the root cause analysis and follow-up with the AI.

What does the AI analyze?​

When you trigger a diagnosis, the AI reads:

  • Recent logs from the affected resource
  • Kubernetes events in the relevant namespace
  • Current infrastructure state (configurations, deployments, dependencies)

It correlates all available signals to surface the root cause, not just the symptom.

Does it apply fixes automatically?​

No. Autodiagnostics is read-only. It surfaces a diagnosis and recommends actions, but all remediation steps require your explicit approval. You stay in control of every change to your infrastructure.

How accurate are the results?​

The AI is most effective for common failure patterns: crashing containers, misconfigured environment variables, resource exhaustion, and dependency connectivity issues. For highly custom setups it provides its best analysis based on available signals — always review recommendations before acting on them in production.