Changelog

v0.0.119

Platform

API keys use the right organization — astroctl and Nova's platform connection now resolve your organization from the API key itself, not from your last interactive astroctl login. A stale local context no longer triggers false "invalid API key" or access-denied errors, so CI automation and Nova-driven workflows stay connected.

v0.0.118

Clusters

Recover a stuck cluster — a cluster could get trapped and refuse to re-register after being moved between organizations or unregistered partway through. Now astroctl infra k8s unregister always frees the cluster so you can register it again, and your cluster is never left in a dead end. Pass --keep-apps to leave your applications running.

Know why a registration failed — when registration is rejected, astroctl infra k8s register status now tells you the real reason instead of sitting on "waiting for agent." If another organization still owns the cluster, it names exactly which organization and cluster to unregister so you can fix it in one step.

Fewer false "unhealthy" alerts — clusters no longer flip to unhealthy on brief, transient blips. The platform now waits for a sustained signal before changing a cluster's status.

v0.0.117

Clusters

High-availability regions — when you create a cluster, only regions with at least three availability zones are offered. Regions with too few zones are filtered out automatically, so every new cluster can spread its control plane and workloads across zones.

More reliable unregister — astroctl infra k8s unregister now removes platform-managed applications before removing the cluster agent, so cleanup completes reliably with nothing left behind. Pass --keep-apps to leave the applications running.

Platform stability — broad reliability hardening so clusters and deployments stay healthy and responsive under heavy load.

Nova in the CLI

Richer investigation output — astroctl nova now renders ranked root causes and workflow steps inline, and shows artifact download links directly in the conversation.

Cleaner messages — internal component and tool names no longer appear in CLI and API output.

v0.0.115

CLI — Registered Clusters (BYOK)

Agent upgrade command — upgrade the agent on a registered cluster without full re-registration:

astroctl infra k8s register upgrade --name <cluster>

Agent version visibility — astroctl infra k8s get <cluster> now shows whether your agent is up to date:

agentVersion: v0.0.101
latestAgentVersion: v0.0.115
updateAvailable: true

Apps cleaned up on unregister — when you unregister a cluster, the platform now asks whether to delete its applications. Pass --keep-apps to skip deletion.

CLI — All Clusters

Template commands — generate a YAML template for any resource type without applying it:

astroctl app template          # application spec
astroctl app profile template  # application profile
astroctl infra k8s template    # cluster spec

Add --schema to see field descriptions, required/optional status, and valid values. Add -ojson for JSON output.

Cloud cleanup hints after delete — after astroctl infra k8s delete, the CLI prints the provider-specific commands to remove IAM roles, service accounts, and WIF pools that remain in your cloud account.

App status exits cleanly — astroctl app status and astroctl app events now exit immediately with a clear error when an application does not exist, instead of reconnecting in a loop.

v0.0.101

⚠️ Breaking Changes — CLI Reorganized

The astroctl command tree has been reorganized. Old commands will not work — update before running any CLI operations.

Old Command	New Command
`astroctl clusters`	`astroctl infra k8s`
`astroctl providers`	`astroctl cloud`
`astroctl org`	`astroctl organization`
`astroctl utils`	Removed
`astroctl availability`	Removed

Update now:

curl -fsSL https://astropulse.io/install.sh | bash
# or
brew upgrade astroctl
# or
astroctl update

Nova AI

Nova Investigation — ask Nova to investigate an incident and it autonomously runs diagnostics, queries your connected tools (Datadog, kubectl, logs), and streams findings in real time. Available in the console and via astroctl nova.

Conversation reset — type /clear in astroctl nova to start a fresh session.

Guardrail messages — when Nova hits a policy limit, it returns a clear explanation of what was blocked and why, instead of a raw error.

Stream cancellation — Nova investigations can now be cancelled mid-stream from the console or with Ctrl+C in the CLI.

Google GKE

Full GKE support — provision, upgrade, and manage Google Kubernetes Engine clusters alongside EKS and AKS:

BYOA with no static credentials required
BYOVPC with regional subnets, secondary IP ranges, and authorized networks
Autopilot mode and release channel support (RAPID, REGULAR, STABLE)
astroctl cloud gcp k8s-versions — list available GKE versions and upgrade paths

See the GKE provisioner docs for full setup and upgrade procedures.

Azure Kubernetes Service (AKS)

Full AKS support — provision, upgrade, and manage Azure clusters with the same workflow as EKS and GKE:

Connect with astroctl cloud azure connect --subscription-id <id> --resource-group <rg> --region <region> --cluster-name <name>
Version upgrades and rolling updates via astroctl infra k8s upgrade
Node group scaling via astroctl infra k8s scale <cluster> <nodegroup> --min/--max
astroctl cloud azure k8s-versions — list available AKS versions

See the AKS provisioner docs for full setup and upgrade procedures.

Self-Hosted Clusters

Cloud connect CLI — set up prerequisites and store credentials for self-hosted clusters on AWS and GCP:

# AWS — automated setup (recommended)
astroctl cloud aws selfHosted setup --account-id <id> --region <region> --cluster-name <name>

# GCP — automated setup (recommended)
astroctl cloud gcp selfHosted setup --project-id <project> --region <region> --cluster-name <name>

See the AWS self-hosted and GCP self-hosted docs for all options including bring-your-own-credentials.

Credential storage — credentials are stored encrypted at rest after setup — no need to provide them on every operation.

Cluster registration — register existing clusters with astroctl infra k8s register.

Role-aware cluster approval — cluster registration requests must be approved by the org owner before the cluster becomes active.

BYOVPC validation — stricter pre-flight checks across all providers catch VPC misconfigurations before provisioning starts.

EKS

Upgrade planning hardened with improved SDK nodegroup operations and dry-run validation
EKS rolling update strategy options expanded
astroctl cloud aws k8s-versions — list available EKS versions and upgrade paths

See the EKS provisioner docs for full upgrade and scaling procedures.

Console

Cluster Upgrade Wizard — guided 5-step upgrade experience (preflight → readiness → checklist → config → confirm) for all providers.

Role-aware cluster approval — cluster registration requests require org owner approval before the cluster becomes active.

Audit Logs — browse organization activity and events directly from the console (Enterprise).

Pod log viewer — multi-container pods now show a container selector so you can view logs from any container without dropping to the CLI.

Nova Investigation view — real-time streaming view of Nova's diagnostic reasoning, tool calls, and findings as the investigation runs.

Platform

Node zone support — workloads can be pinned to specific availability zones across all provisioners (EKS, GKE, AKS, self-hosted).

Named API keys — API keys now support an optional name field.

Installation

curl -fsSL https://astropulse.io/install.sh | bash

Or via Homebrew: brew upgrade astroctl | Self-update: astroctl update

March 2026 Release Highlights

Nova

Nova Cloud gives teams a browser-based Nova workspace with shareable conversations and a managed starting point.
Nova Connect is the hosted remote MCP endpoint at https://platform.astropulse.io/mcp for supported editor and CLI integrations such as Claude Code, Cursor, VS Code, and Claude Desktop.
Nova Direct remains the self-hosted option for teams that want local Docker-based execution and the built-in MCP Marketplace.
astroctl nova provides terminal-native Nova sessions with resume and continue workflows.
Slack-connected workflows support approvals, notifications, and team visibility around Nova activity.

Bring Your Own Kubernetes

Register existing EKS, GKE, AKS, on-prem, and local Kubernetes clusters with astroctl infra k8s register.
Use --no-install for split-team handoffs and --dry-run to preview what will be deployed before applying it.
Registered clusters use the same deployment, health, and console workflows as platform-provisioned clusters.

Platform Experience

Docs now clearly separate Nova Connect (hosted) and Nova Direct (self-hosted) so setup paths are unambiguous.
Editor and CLI setup centers on the hosted OAuth sign-in flow for Nova Connect.

v0.0.98

Nova AI — One AI, Any Interface

Nova is a skill-based AI platform engineer. The same AI, the same skills — across every interface:

Nova Cloud — AI assistant in the browser, shareable conversations, zero setup
astroctl nova — Terminal-native chat with rich rendering, session resume, paste support
Slack — @mention Nova in any channel for team collaboration and async workflows

Nova ships with built-in skills: Cloud Cost Calculator, Slack, GitHub, and Infrastructure Tools. Skills are extensible — enable them from the Skills panel or add custom ones via MCP Marketplace.

New CLI Commands

astroctl audit — View audit logs and organization activity (Enterprise)
- astroctl audit list — List audit events
- astroctl audit get — Get event details
- astroctl audit export — Export events to JSON or CSV
- astroctl audit stats — Audit statistics (Admin/Owner)
astroctl update — Check for updates and install the latest CLI version
astroctl application profile tiers — List available profile tiers

CLI Breaking Changes

The CLI command tree has been reorganized for consistency:

Old Command	New Command
`astroctl clusters`	`astroctl infra k8s`
`astroctl providers`	`astroctl cloud`
`astroctl org`	`astroctl organization`
`astroctl utils`	Removed
`astroctl availability`	Removed

Console

Cluster Upgrade Wizard — Guided 5-step upgrade experience in the console
Audit Logs — Browse organization activity from the web console (Enterprise)

Installation

curl -fsSL https://astropulse.io/install.sh | bash

Or via Homebrew: brew upgrade astroctl | Self-update: astroctl update

v0.0.97

New Features

Nova Managed Cloud: AI-powered cloud infrastructure assistant
- Natural language cluster provisioning and management
- Interactive workflows for deployments, upgrades, and troubleshooting
- Real-time cluster analysis and recommendations
- Integrated with all provisioners (EKS, GKE, Self-Hosted)
Google GKE Provisioner: Full support for Google Kubernetes Engine (GKE)
- BYOA (Bring Your Own Account) with no static credentials required
- BYOVPC with regional subnets and secondary IP ranges for pods/services
- Autopilot mode for fully managed node operations
- Release channel support (RAPID, REGULAR, STABLE)
- Authorized networks for API server access control
- Private clusters supported

Cluster Updates & Upgrades

Console Upgrade Wizard: New 5-step guided upgrade experience for all provisioners
1. Preflight Check - Validates upgrade paths and checks for issues
2. Readiness Report - Comprehensive upgrade analysis with risk score
3. Pre-Upgrade Checklist - Interactive checklist with verification commands
4. Configuration - Select target version and rolling update settings
5. Confirmation - Review summary and confirm by typing cluster name
CLI Commands: Simplified cluster management commands
- astroctl infra k8s upgrade <cluster> <version> - Version upgrades with dry-run and readiness reports
- astroctl infra k8s scale <cluster> <nodegroup> --min/--max - Quick node group scaling
- astroctl infra k8s update <cluster> -f update.yaml - Advanced YAML-based updates
- astroctl infra k8s progress stream/get/history - Real-time progress monitoring
Version Discovery: New commands to check available Kubernetes versions
- astroctl cloud aws k8s-versions - EKS versions with upgrade paths
- astroctl cloud gcp k8s-versions - GKE versions with upgrade paths
- astroctl cloud aws/gcp k8s-versions --self-hosted - Self-hosted version support

Provisioner Updates Support

Provisioner	Version Upgrade	Node Scaling	Rolling Updates	Progress Monitoring
AWS EKS	✅	✅	✅	✅
Google GKE	✅	✅	✅	✅
AWS Self-Hosted	✅	✅	✅	✅
GCP Self-Hosted	✅	✅	✅	✅

Provider-Specific Update Options

EKS: AMI family selection (AmazonLinux2023, Bottlerocket, Ubuntu)
GKE: Image type, release channel, maintenance windows, authorized networks
Self-Hosted: Control plane configuration and node group management

Breaking Changes

CLI Upgrade Required: Older CLI versions (< v0.0.97) are not compatible with this release
Install the latest CLI: curl -fsSL https://astropulse.io/install.sh | bash
Or via Homebrew: brew upgrade astroctl
Existing cluster configurations remain compatible - no changes needed to your cluster specs

v0.0.90

Bug Fixes

astroctl: Fixed log reading from pods with multiple containers - now properly supports selecting and reading logs from specific containers
Console: Fixed container log viewer to handle multiple containers - users can now select which container's logs to view in the web console

Improvements

Enhanced container selection UI in the Console's log viewer
Better error handling for multi-container pod scenarios

v0.0.84

We are excited to announce the first public release of AstroPulse Platform - a GitOps platform for cost-efficient application deployment across public cloud infrastructure.

✨ Key Features

Application Management: Deploy and manage applications with GitOps best practices
Multi-Cloud Clusters: Provision and manage Kubernetes clusters - AWS EKS (managed), self-hosted K8s on AWS and GCP
Real-time Events: Stream application and cluster events with WebSocket support
Notifications: Integrated Slack, email, and webhook notifications
Cost Optimization: Built-in Karpenter support for intelligent autoscaling
CLI Tool: Comprehensive astroctl command-line interface

AstroCtl CLI

** Other Platforms:** Visit the Get Started guide

📖 Documentation

Full documentation available at AstroPulse Docs

🙏 Thank You

Thank you to our users, customers, and launch partners for making this release possible!

Get Started: Get Started →

v0.0.119​

Platform​

v0.0.118​

Clusters​

v0.0.117​

Clusters​

Nova in the CLI​

v0.0.115​

CLI — Registered Clusters (BYOK)​

CLI — All Clusters​

v0.0.101​

⚠️ Breaking Changes — CLI Reorganized​

Nova AI​

Google GKE​

Azure Kubernetes Service (AKS)​

Self-Hosted Clusters​

EKS​

Console​

Platform​

Installation​

March 2026 Release Highlights​

Nova​

Bring Your Own Kubernetes​

Platform Experience​

v0.0.98​

Nova AI — One AI, Any Interface​

New CLI Commands​

CLI Breaking Changes​

Console​

Installation​

v0.0.97​

New Features​

Cluster Updates & Upgrades​

Provisioner Updates Support​

Provider-Specific Update Options​

Breaking Changes​

v0.0.90​

Bug Fixes​

Improvements​

v0.0.84​

✨ Key Features​

📖 Documentation​

🙏 Thank You​

v0.0.119

Platform

v0.0.118

Clusters

v0.0.117

Clusters

Nova in the CLI

v0.0.115

CLI — Registered Clusters (BYOK)

CLI — All Clusters

v0.0.101

⚠️ Breaking Changes — CLI Reorganized

Nova AI

Google GKE

Azure Kubernetes Service (AKS)

Self-Hosted Clusters

EKS

Console

Platform

Installation

March 2026 Release Highlights

Nova

Bring Your Own Kubernetes

Platform Experience

v0.0.98

Nova AI — One AI, Any Interface

New CLI Commands

CLI Breaking Changes

Console

Installation

v0.0.97

New Features

Cluster Updates & Upgrades

Provisioner Updates Support

Provider-Specific Update Options

Breaking Changes

v0.0.90

Bug Fixes

Improvements

v0.0.84

✨ Key Features

📖 Documentation

🙏 Thank You