v0.1.0

Changelog

• 7a7474f feat: check engine core - Finding, Check interface, shared helpers
• 65f8076 feat: cost model with instance pricing table and per-resource fallback
• 6b52d6e feat: idle-gpu check - unrequested GPU nodes priced at full node cost
• 66a54f9 feat: idle-loadbalancer check
• 577c657 feat: kubeoptimizer scan CLI with output/rate flags
• 42741d5 feat: no-requests check - warn on workloads without resource requests
• 249c5d6 feat: overprovisioned-requests rough right-sizing via metrics-server
• 27a9930 feat: release automation with goreleaser (brew cask + krew manifest)
• 5c442b3 feat: report aggregation with table and JSON renderers
• 92b22e8 feat: snapshot collector with per-resource failure isolation
• 2d66e60 feat: underutilized-nodes check - sub-50% request nodes as consolidation candidates
• 99c4704 feat: unused-pv check - released/available PVs and orphan PVCs
• e3a974e feat: zombie-workloads check - crashloop cost, stale finished pods/jobs
• 0a23d92 fix: adapt metrics fake fixture in collector test
• 9f42246 fix: clamp right-sizing recommendation to current request (negative savings bug)
• 8be556d fix: isolate e2e kind kubeconfig from user's ~/.kube/config
• 0910b99 fix: pass true nil metrics interface on construction failure; tidy go.mod
• e6021bc fix: skip idle GPU nodes in underutilized-nodes check (double-count)
• 4020cd9 fix: surface non-NotFound metrics errors instead of masking as absence
• 9e57db3 refactor: rename module path to github.com/DPS0340/kubeoptimizer