Argus
Know what you run, whether it is still safe and supported, and which version to upgrade to.
Upstream projects ship releases every day. Argus discovers the third-party software in your clusters and filters that stream down to what affects you, with the evidence behind every answer.
Read-only access to your clusters. It installs nothing, changes nothing, sends nothing out.
Sound familiar?
A cluster runs dozens of third-party products. Each has its own release cadence, support window and compatibility rules, and none of them will tell you.
End of support comes as a surprise
You usually learn that a product is out of support during an incident or an audit. By then the upgrade is no longer a plan, it is an emergency.
"Does this affect us?" takes hours
When an advisory lands, finding out which version runs in which cluster takes longer than reading the advisory.
Kubernetes does not wait
Kubernetes ships three releases a year. Managed services either force-upgrade old versions or charge extra for extended support.
The order of upgrades is unclear
Cluster first, or application first? Done in the wrong order, an upgrade leaves you on a combination nobody supports.
The information is scattered
Release notes, support pages, compatibility matrices and advisories live in dozens of places. A hand-kept list is stale the day it is written.
Management cannot see the whole
Teams know the details; management wants risk, cost and priority. Nothing lets both talk from the same data.
From "17 updates available" to the two that matter today
Discovery and recognition
Reads workloads, images and Helm releases, and matches them to products, through mirrors and proxy registries. Anything unrecognized is listed, never guessed.
Recommended, not just latest
The smallest upgrade that keeps you supported: the newest patch of a supported line, or the oldest line supported for 90 more days.
Kubernetes upgrade readiness
READY or NOT READY for the next minor version, listing every installed line the vendor does not support on it, and the deprecated APIs still in use.
Ordered upgrade plan
Every upgrade in an order that keeps each step on supported combinations, including vendors that forbid skipping versions.
Evidence, not opinions
Each answer cites its source: the label that gave a version, the vendor matrix that excluded a candidate, the page a support date came from and when.
Timeline, summary and alerts
A Kubernetes support calendar across clusters, a printable management summary, and scheduled email reports per organization or cluster.
A deterministic core: same cluster, same data, same answer
Discover
Read workloads, images and Helm labels through read-only list calls.
Recognize
Match images to the curated product catalog by rule, never by guess.
Inventory
Group components into installations, recording evidence for each version.
Assess
Check releases, support calendars, advisories and vendor compatibility.
The latest release is not always the right target
Argus recommends the smallest step that keeps you supported on your Kubernetes version. For Kubernetes itself it gives the next minor version and the full path to a supported one. For a retired product, it recommends the successor.
160+ products, curated and growing daily
Beyond the catalog, Helm releases are matched to their charts and every other image is compared with its registry's tags. Your own images are listed separately and never assessed.
Not a scanner. Lifecycle intelligence.
Vulnerability scanners look inside the image: packages, layers, CVEs. Argus works one level up: it recognizes products and says "this version of this product left support last October". It does not replace your scanner; it completes it.
It never guesses
It does not pass off software it cannot recognize as recognized. What it does not know, it calls unknown.
It only reads
It installs nothing in your cluster, changes nothing, runs nothing. It reads no Secrets. It is one small service.
Your data stays with you
Your clusters' inventory stays in your own installation. Argus only fetches reference data; air-gapped sites load it by hand.
Fresh data every day
Releases, support calendars, advisories and compatibility tables are updated, audited and distributed signed, every day.
Many clusters, one view
Every cluster and organization in one place. Each team hears only about its own clusters.
First results in minutes
Installation is one Helm command. The first assessment is on screen minutes later, the management summary on the table the same day.
Three steps to everyday use
Introduction and demo
In a half-hour call we listen to how your environment looks and show Argus live.
Pilot
Installation on one cluster, the first assessment and management summary the same day. Read access to the cluster is all it needs.
Roll-out
Further clusters are added, notifications are set per team, and an upgrade calendar takes shape.
Frequently asked
Does it install or change anything in my cluster?
No. Argus observes your clusters with read permissions only; it changes no resource, runs no command and reads no Secrets. You do the upgrading; Argus tells you when, to what, and in which order.
Which environments does it support?
Kubernetes itself and distributions such as EKS, AKS, GKE and RKE2, on premises and in the cloud. One Argus can watch many clusters.
Which software does it recognize?
More than 160 products common in the Kubernetes ecosystem. You can define your own products too. Even for images it does not recognize, it tells you whether a newer version exists.
Does it replace our security tools?
No. Vulnerability scanners look at the packages inside images; Argus looks at products' versions, support status and upgrade paths. The two work side by side.
Where does our data go?
Nowhere. Your clusters' inventory stays in your own Argus installation. Argus only downloads reference data, and in air-gapped environments that data can be loaded by hand.
How is it priced?
We prepare a quote for your scope. Write to us and let us talk.
Ask for a scan of your clusters
A scan, a demo or a technical question: tell us a little and the right team answers. A technical overview for your engineers is ready on request.
Reach us directly
- info@composiv.ai
- Phone
- +90 212 328 08 25
- Office
- ITU Ayazağa Teknokent, ARI-3
Maslak, Istanbul, Türkiye