Roadmap

Build the investigation path first. Earn the integrations after.

This is direction, not a release calendar. Real cluster evidence can reorder the work; shipped capabilities belong on the product pages, not under “coming soon.”

01
Now · v0.3

Evidence-first baseline

The local incident workflow available for use against a real cluster.

  • —Local Go process with an embedded web UI
  • —Existing kubeconfig, context selection, and read-only doctor preflight
  • —Evidence-backed incident brief with deterministic local rules
  • —Explicit separation of observed facts, hypotheses, caveats, and next check
  • —Namespace environments and workload drill-down
  • —Pods, readiness, restarts, reasons, events, and logs
  • —Node inventory, capacity, placement, and optional metrics
  • —Karpenter NodePool provisioned capacity against configured limits
  • —Optional ingress-nginx Prometheus adapter
  • —Synthetic incident demo and cross-platform release archives
02
Next

Evaluation and hardening

Work driven by labeled incidents and real cluster behavior, not a fixed quarterly promise.

  • —Validation across RBAC and authentication setups
  • —A versioned Kubernetes incident evaluation set
  • —Measure helpfulness, false certainty, unknown rate, latency and cost
  • —Local history with documented retention
  • —More explicit freshness and source diagnostics
  • —Additional traffic source validation
03
Later

Decision expansion

Useful directions that need evidence before they become commitments.

  • —Optional remote decision models only after evaluation gates pass
  • —Additional ingress-controller adapters
  • —Cloud pricing and billing enrichment beyond the current capacity view
  • —Optional database summaries
  • —Namespace grouping beyond the one-namespace-one-environment model
  • —Cross-environment image digest comparison

Need a shared cluster workflow?

Pro covers an individual fleet up to 10 contexts. Talk to an engineer when the requirement is shared access, a larger fleet, or an integration.

Talk to an engineer →

Try the current product on your machine.

Install with your coding agent, start with synthetic data, then select a kubeconfig context you control.

or install yourself

Local install · read-only first · no Helm chart