Changelog
What we shipped.
Kubeadapt changes every day. Here is how much, and what it means for you.
How much we shipped
Yesterday, November 20
- changes shipped
- 23
- lines added and removed
- +58,927 −14,380
- parts of the product touched
- 3
Every day for the last 12 months
496 changes in the last 30 days. Updated every hour.
Vote on what we build nextWeek of November 16
The changes you will notice.
- New
Reports is open to everyone in your organization.
Choose the executive or engineering format, open each report on its own page with its history, and see every team's cost in both.
- New
Every persistent volume has its own page.
See a volume's run rate, size, price per GiB-month and daily cost.
- New
Developers has its own place in the sidebar.
See your API keys with their scopes and last use, and copy curl and CLI commands to make your first API call.
- Improved
Jobs and CronJobs are priced by the minute they run.
Batch workloads show what their runs cost over the last 30 days instead of a $0 run rate between runs.
- Improved
Cost trend charts keep batch costs in their own lane.
With per-minute pricing, the cluster and organization cost trends still show what batch workloads cost under Batch.
- Improved
Settings gives each section its own page.
Members has search, filters and paging for large teams, and single sign-on walks you through setup for Okta, Microsoft Entra ID and Google Workspace.
- Improved
Cluster preferences show when the agent lost its connection.
A seven-day hourly grid lists every interruption, and on-premises pricing now adds back to exactly the bill you enter.
- Improved
Saved views in Cost Explorer stay open after a reload.
The view you opened is kept in the link, so a reload or a shared link keeps it, and saving updates it instead of making a copy.
- Improved
Cost Explorer marks run rates that cover only part of a group.
A grouped row whose run rate adds up only the parts with a current rate reads ≥ $X/mo, in the table and in exports.
- Fixed
Daily averages start from the first day with recorded cost.
Days before a cluster recorded any cost no longer count as $0 days in Cost Explorer's daily average.
- Fixed
The CLI keeps missing cost amounts empty in JSON and YAML output.
When there is no current figure, scripts now read null instead of an empty amount that broke number parsing.
Releases
18 entries
2026
13 entries
- feat(2mo ago)
Kubeadapt MCP: Ask Your AI Tools About Your Kubernetes Costs
Connect Claude, ChatGPT, Cursor, VS Code and other AI tools to Kubeadapt and ask about spend, idle capacity, rightsizing and alerts. Read-only.
- mcp
- ai
- integrations
- developer-experience
- new-feature
- feat(2mo ago)
FOCUS Cost Export and Helm Chart 1.0
Export Kubernetes cost as a FOCUS 1.4 file from Cost Explorer or the public API. Plus Helm chart 1.0, which renames the eBPF agent subchart.
- export
- focus
- api
- helm
- agent
- feat(8mo ago)
Smart Alerting: Catch Cost Issues Before They Hit the Bill
Update, Sep 23, 2026: Corrected to match what ships. Cost-spike detection is now described as it works today. Quiet hours, timed mutes, unused-resources alerts and notify-only optimization alerts were removed from this entry because they are not available.
Rules, channels and notification policies that watch your spend: cost-spike detection fitted to each scope's history, budget alerts and new-workload alerts.
- alerting
- cost-optimization
- incidents
- slack
- new-feature
- feat(8mo ago)
Kubeadapt AI: Plain-Language Incident Explanations
Update, Sep 23, 2026: Corrected. An earlier version said utilization metrics, cost figures and workload details never reach the model. They do. The section below lists what is sent.
Kubeadapt AI explains alert incidents in plain language. It is off by default, and this entry lists exactly what data the model sees.
- ai
- privacy
- incidents
- cost-optimization
- new-feature
- improve(8mo ago)
Accurate Pricing for Azure Files Premium Volumes
Shared (ReadWriteMany) PVs on Azure Files Premium now get a real price instead of a wrong static rate. Volumes are matched to file-share prices by CSI driver.
- azure
- storage
- pricing
- cost-accuracy
- feat(8mo ago)
Cost Attribution Rules: Assign New Workloads Automatically
Define rules once and Cost Attribution assigns new workloads, namespaces and volumes to the right team or department as they appear.
- cost-allocation
- teams
- automation
- new-feature
- feat(8mo ago)
Automatic Updates for the Kubeadapt Agent
A sidecar that checks for safe chart versions, runs version upgrades, and reports the result back to the dashboard.
- agent
- kubernetes
- helm
- automation
- infrastructure
- feat(8mo ago)
kubeadapt-cli: Your Cluster Costs from the Terminal
A CLI to query clusters, costs, recommendations and workloads from your terminal. Install with Homebrew or download the binary. Table, JSON and YAML output.
- cli
- developer-experience
- new-feature
- improve(9mo ago)
Introducing kubeadapt-agent: Lightweight Kubernetes Metrics Collector, Rebuilt from Scratch
The original agent is deprecated. kubeadapt-agent drops the Prometheus stack, needs only metrics-server, and auto-discovers GPU workloads.
- agent
- kubernetes
- gpu
- performance
- breaking-change
- feat(9mo ago)
AWS Savings Plans & Reserved Instance Attribution
Connect your AWS account and see RI/SP discounts reflected in your actual cluster costs instead of on-demand rates.
- aws
- cost-optimization
- savings
- feat(9mo ago)
Teams & Departments: Know Who's Spending What
Group costs by team and department using Kubernetes labels or dashboard assignment. See per-team spend, department rollups, and cost trends.
- teams
- cost-allocation
- dashboard
- new-feature
- feat(9mo ago)
Public API and API Key Management
Create API keys from Settings, authenticate with a bearer token, and query clusters, costs, and recommendations via REST.
- api
- developer-experience
- new-feature
- feat(9mo ago)
Feedback Portal: Public Roadmap and Feature Requests
See what we're building, vote on what matters to you, and submit feature requests directly. Our roadmap is public now.
- community
- roadmap
- new-feature
2025
5 entries
- feat(12mo ago)
Capacity Planning: Right-Size Your Infrastructure
Visualize node group utilization, identify oversized infrastructure, and run what-if scenarios for instance type changes, spot ratios, and Graviton migration.
- infrastructure
- cost-optimization
- nodes
- feat(1y ago)
Best Practices & Configuration Scoring
Audit Kubernetes configurations against 50+ best-practice rules. Get a score per cluster and prioritized fixes for security, reliability and resources.
- best-practices
- security
- reliability
- feat(1y ago)
Spot Instance Migration Recommendations
See which workloads can safely run on spot instances and what they would save. StatefulSets, PVs and system-critical pods are filtered out.
- spot-instances
- cost-optimization
- infrastructure
- feat(1y ago)
Orphaned Resources Cleanup
Find PersistentVolumes that are no longer attached to any workload and see how much they cost. Clean up storage waste you didn't know existed.
- storage
- cost-optimization
- cleanup
- feat(1y ago)
Abandoned Workloads Detection
Update, Sep 11, 2026: Disabled on 2026-09-11 while we fix false positives. Low CPU and memory usage alone flagged services that were still in use. We will bring detection back once it can also use ownership and traffic signals. Orphaned PersistentVolume detection is unaffected.
Find workloads with zero or near-zero utilization sitting idle in your cluster. See what they cost and decide whether to scale down or remove them.
- workloads
- cost-optimization
- automation