Show HN: Debug Kubernetes using an infra context graph (opens in new tab)

Hey! We just released Anyshift CLI. We’ve been working on a CLI that helps debug production issues by adding infrastructure context to logs.

At Anyshift, we build the context layer agent in production. To do so we build a versioned graph of production infrastructure across AWS, Kubernetes, Datadog, GitHub, and more. Once connected, our CLI allows to debug incidents or perform deep analysis of what's happening or what happened in your prod.

For example:

kubectl logs -n rbl -l app=backend --tail=100 | anyshift ask "what's wrong?"

The scenario here: A Kubernetes backend is returning 502s every 5 seconds on /status. Logs show service discovery failures, but nothing obvious stands out. Instead of scanning manually you could pipe the logs into Anyshift.

It analyzes them against a pre-built dependency graph of your infrastructure: every resource, connection, and change over time.

Then it returns:

- failure chain (Redis cache miss → missing SERVICE_REGISTRY_URL → EC2 API resolution failure) - Impact scope (only external monitoring affected, no user traffic) - fix

Happy to answer questions!