major release · October 13, 2026

---

🚀 New Features​

• Import Existing AWS (beta)
  SleakOps can now adopt your existing infrastructure — AWS accounts, networks, EKS clusters, domains, databases and other dependencies, projects and their variable groups — instead of provisioning everything from scratch. Nothing gets recreated: imported resources are flagged as such, databases can be onboarded by pasting their connection string, and Valkey is supported alongside the other managed dependencies.

• Static Sites
  A new static project type builds your frontend from Git and serves it on your own domain through S3 and CloudFront, with no container or web server to maintain.

• Manifests (beta)
  The manifests SleakOps generates for your workloads are now visible and editable, and you can add your own (for example, an extra Ingress) for full transparency over what gets deployed. NodePools also gain an extended form with YAML editing, so advanced Karpenter settings no longer have to be changed directly in the cluster.

• Environment Presets (beta)
  Preset details now show every project at once and can create all of them, with their resources, in a few clicks instead of one project at a time. Presets exported locally also work with private repositories.

• MCP Server (beta)
  Connect the SleakOps MCP server to Claude Code or any MCP client and operate the platform in natural language.

• Contextual support chat
  Every resource now has a support button that opens the support chat directly from where you are working.

---

✨ Improvements​

• Essential addons managed from the Addons section
  Essential cluster addons are now listed and managed together with the rest of the addons. Their configuration is protected from edits that could break the cluster, and it is completed automatically when they are adopted.

• Clearer Build and Deployment status
  Builds and Deployments now show why they are waiting (for example, queued behind another build or waiting for a resource to be ready), and pending changes are reviewed directly from the deployment detail.

• Console and access improvements
  Advanced settings open automatically when they differ from the defaults, addon links are available from the Dashboard and Workloads, Settings has its own sidebar, and services show all of their variable groups. In addition, the Viewer role can no longer read the secret values of variable groups.

---

🐛 Bug Fixes​

• Safer database dependency edits
  Saving a database dependency no longer disables its deletion protection in AWS, Aurora clusters restored from a snapshot or point-in-time can be edited again, and the DocumentDB edit form shows the configured instance class. Restores now validate the engine version against the snapshot before creating any resources.

• Reliable environment domain changes
  Changing an environment's domain now provisions the new domain, re-points addons and S3 aliases, and redeploys affected projects (including static sites and workers) only after the domain is ready. Orphaned domains can be deleted without errors.

• Cluster creation and deletion
  New clusters install their essential addons in the correct order, and deleting a cluster no longer fails when some addons were already being removed or leaves project namespaces behind.

• Workloads and activity log
  Re-running a CronJob now opens the job form prefilled with its definition, service updates keep their backup replica settings, removing a file-type variable group no longer leaves pods trying to mount it, and the Dashboard activity log shows only the selected provider's activity.

---

minor release · July 16, 2026

---

🚀 New Features​

• Addon and Environment Import/Export
  Adopt Helm releases already running in your cluster as SleakOps Addons, create custom Addons from JSON, and export or import the resources of a project or environment.

• Accounts & Networking Explorer
  A new section, opened from the Providers list, shows each Account of a Provider and a read-only view of its networking (VPC, subnets, NAT and Internet gateways).

---

✨ Improvements​

• New Webservices Form for Everyone
  The redesigned webservices form is now enabled for all users and replaces the previous one.

• Simpler Account Selection
  The sidebar lists only accounts that have at least one cluster and auto-selects the account when only one is available.

• Improved AI Reports
  Infrastructure diagrams now render instead of showing raw code, every report can be printed to PDF, and all reports share a consistent layout.

• Redis Cluster Monitoring
  Metrics now work for Redis dependencies in cluster mode, with a tab for each member of the replication group.

• EFS Addon Update
  A new version of the EFS Addon with an updated CSI driver is available, and file systems with associated volumes are protected against changes that would break them.

---

🐛 Bug Fixes​

• Volume Deletion
  Deleting a volume still mounted by a pod now retries until it is released, or ends in a clear error instead of failing silently.

• Dependency Creation Errors
  The real validation error is now shown when creating a dependency fails, instead of a generic message.

• Dockertron Presets
  Fixed worker commands, dependency variable names, and n8n settings in Dockertron presets.

• Billing Forecast Comparison
  Cost summary cards now compare the forecast against the full previous-month total, fixing inflated percentages.

• Mobile Dashboard Font Size
  Fixed text that was too small on the dashboard on mobile devices.

---

patch release · July 6, 2026

---

✨ Improvements​

• Sticky Sessions Tutorial for Web Services
  New guide on enabling sticky sessions in a Web Service through its ingress annotations, including when stickiness is the wrong tool and how to verify it.

• Cost Review Guide
  New tutorial on reading your AWS spend with the Cost Explorer, Kubecost, and Cost Allocation Tags before deciding what to optimize.

• Support Scope Page
  New page explaining how to reach Support, what it covers, and where to look first for common questions, aligned with the Responsibility Model.

• Exhaustive Environment Testing Guide
  New checklist of tests to validate that your application is fully functional after migrating to a new SleakOps infrastructure.

• Database Encryption Documentation
  The dependency docs now accurately describe at-rest encryption defaults, opt-out, and custom KMS key options for database dependencies.

---

minor release · June 30, 2026

---

🚀 New Features​

• Kubernetes 1.34 Support
  Clusters can now be upgraded from EKS 1.33 to 1.34, and new clusters are created on 1.34 by default. Managed addons such as Prometheus and the AWS Load Balancer Controller were updated to the APIs recommended for this version.

• Compatibility Report for Self-Managed Manifests
  Before an EKS upgrade, you can generate a report that flags resources deployed outside SleakOps that use APIs deprecated or removed in the target Kubernetes version.

• AWS DataSync for S3 Dependencies
  A new action on S3 dependencies syncs data from another bucket or an external source using AWS DataSync, with transfer status visible from the console.

---

✨ Improvements​

• Cluster Upgrades in the Upgrades Section
  EKS upgrades are now scheduled, executed and tracked from the Upgrades section, with email confirmation, a 24-hour reminder and a per-service downtime report. Clusters with a pending upgrade are flagged in the cluster list, and nodes rotate only once during the upgrade.

• Automatic Timeout for Stuck Builds
  Builds that exceed the maximum allowed time are now stopped automatically and marked as failed, without depending on the CLI or CI staying connected.

• Interface Refinements
  Redesigned list fields in forms, new date pickers, better positioning of selectors inside drawers and visual fixes to the Explorer.

---

🐛 Bug Fixes​

• RDS Multi-AZ Disabled by Default
  New RDS databases no longer enable Multi-AZ unless you request it, avoiding unexpected costs.

• Form Fixes
  The web services form now keeps saved healthcheck values when editing, and the variable group and chart dependency forms no longer show incorrect validations.

• Deployments After Changes Without Deploy
  Accepting a pending deployment after editing a workload without deploying no longer fails.

• Project Access Updates
  Policy changes in Project Access are now applied and reflected in the environment status.

• Port Visible Without Healthcheck
  The service detail shows the port even when the healthcheck is disabled.

• Environment Deletion Confirmation
  The confirmation dialog no longer lists internal billing objects.

• Chat Input After Refresh
  The chat input no longer gets blocked when the conversation refreshes.

---

minor release · June 11, 2026

---

🚀 New Features​

• Expanded RDS and Aurora Management
  RDS dependencies can now be restored to a specific point in time and use custom parameter groups. Storage encryption is enabled by default, with custom KMS keys, Performance Insights and snapshot restore for Aurora PostgreSQL.

• Kafka Authentication Options
  Kafka (MSK) dependencies can now be created with IAM or SCRAM (username/password) authentication, in addition to no client authentication.

• Multiple EFS File Systems per Cluster
  EFS is now managed as a cluster addon and supports more than one file system per cluster. Clusters with legacy storage classes are migrated automatically to the current ones.

• AI Cost Breakdown by Product
  AI spending history is now broken down by product (chat, autodiagnostics, reports, etc.). Report requests check the AI budget limit up front and show a link to adjust it when reached.

---

✨ Improvements​

• Resource Details Everywhere
  Resource details can be opened from any screen and expose the same actions available in tables. Migration completion notifications link directly to their detail.

• Billing Screen Improvements
  Billing loads faster, groups costs by resource type and shows costs per nodepool in each cluster's detail.

• BuildKit Migration for Production Projects
  Production projects still using Kaniko are notified in advance and then migrated to BuildKit.

• More Reliable Log Compaction
  The Loki and LokiV2 addons now provide persistent storage to their compaction component, so it runs without errors.

---

🐛 Bug Fixes​

• Activity Log Accuracy
  Volume and workload actions now produce a single entry on the correct resource, and project clones are recorded and can be filtered.

• SSL Certificate Validation
  Validating an environment domain's SSL certificate now consistently creates the deployment pending approval.

• Spellcheck Disabled on Sensitive Fields
  Fields holding secrets (variable values, API keys, tokens, kubeconfig, AWS keys) no longer send their content to the browser's spellchecker.

• S3 Bucket Name
  S3 dependencies now always use the actual bucket name instead of assuming it matches the dependency name.

• Hidden Table Filters
  Hidden filters are now kept in the URL, so filtered table views survive a refresh and can be shared.

---