Kakveda open source-level AI, Infra observability agent

Over the last 48 hours we shipped a major update to our open-source observability + infra + AI runtime stack.

Big changes:

Renamed kakveda-aankh → kakveda-netra

Built as a real host-level agent (CLI + systemd support)

Background mode (--start, --stop, --bg-status)

Full infra metrics:

CPU, memory, disk, network, load

Docker container metrics

temperature, FD usage, diagnostics

Kubernetes inventory:

nodes, pods, deployments, services, configmaps, secrets

kubeconfig auto-detection

Dashboard upgrades:

golden signals

trend charts

thresholds + alert coloring

detailed drilldowns

Custom dashboards:

saved layouts

YAML import/export

Runtime control of host agent from dashboard

Cost comparison docs (Datadog, AppDynamics, Logz.io, LangSmith etc.)

The idea is simple: Infra monitoring + AI/agent governance + observability in one platform.

Still early. Would love feedback from infra engineers and K8s folks. Github : https://github.com/prateekdevisingh/kakveda