Natan Yellin

Fix bug with prometheus embeds

2

Sometimes holmes tries to create an embed based on Prometheus instant queries, which is wrong. It should only create embeds based on range queries.

Less server errors when not connected to saas

1

I've tested in the case where SaaS is not available (i.e. no token) - please also test on happy path where it is available and verify everything still works.

Alpha: allow running proactive checks with HolmesGPT

2

Adds Kubernetes operator for HolmesGPT with two CRDs: * HealthCheck CRD - a single human-language health check (pass/fail criteria) that optionally sends a notification on failure * ScheduledHealthCheck - creates...

### What would you like to be added? Allow running HolmesGPT in the background in your cluster ### Why is this needed? Move from reactive to proactive monitoring https://github.com/robusta-dev/holmesgpt/pull/836

Speed up holmes by limiting data fetched from kubectl logs

2

When there are many logs, it takes a while to run `kubectl logs` and we throw away most of the data anyway due to token limits: We should limit the...

### What would you like to be added? This way you don't need to configure explicit access ### Why is this needed? Requested by @prashanthm10