AlertMonitor Feature

Security Validation

Every system. Every night. The nightly fleet scan inventories the software on every device and correlates it against vulnerability data — apps, custom apps, CMS components, and the language ecosystems your developers pull from.

It finds the software package managers do not know about, flags what is actively exploited, hunts for supply-chain compromise, and turns what it finds into remediation jobs your team approves.

Security Validation — Nightly Fleet Scan
InventoryCorrelateSupply chainRemediate

Enumerating every package manager on every device

Fleet0/28 devices
Sources discovered
apt pkgs
rpm pkgs
npm-pkg pkgs
pip-venv pkgs
gem pkgs
nuget pkgs
manual pkgs
Packages
0
CVEs
0
Flagged
0
Jobs
0

The dashboard

The whole fleet, scored by what actually matters

Raw CVE counts are noise. Security Validation separates the total from the four that are critical, the one that is on CISA's actively-exploited list, and the twenty-five sitting on internet-facing components. That is the queue you work.

AlertMonitor — Security Validation
Security Validation
Software vulnerabilities & zero-days across apps, custom apps, CMS & supply-chain — from the nightly fleet scan.
Scan All DevicesRetestDeveloper SystemsRefreshCSVPDF Report
FindingsRemediationSupply ChainAsk AIHistory
Total CVEs
176
Critical
4
High
69
Actively Exploited
1
Public-Facing
25
1 actively exploited.A CVE in CISA's KEV catalog outranks a higher CVSS score nobody is using.

Shown in the portal's dark theme. AlertMonitor ships with both dark and light themes — pick per user.

Coverage

Every package manager. Every language. Every device.

Most vulnerability tools ask the package manager what is installed and stop there. That answer is incomplete on every real fleet. Security Validation collects from the package databases and from the filesystem, so the inventory includes the software nobody registered.

OS Packages

The system package databases, read directly on the device — no credentialed remote scan, no scan window.

apt / dpkgDebian, Ubuntu
rpmRHEL, Rocky, Alma
yumLegacy RHEL / CentOS
dnfFedora, RHEL 8+
zypperSUSE, openSUSE
snapConfined packages
flatpakDesktop apps

Language Ecosystems

Where modern risk actually lives. Dependencies are inventoried per interpreter and per project, not just globally.

pipSystem Python
pip-venvPer-project virtualenvs
npmGlobal Node packages
npm-pkgPer-project node_modules
gemRubyGems
composerPHP / CMS plugins
nuget.NET on Windows

Unmanaged Software

Software no package manager knows about. This is the category conventional scanners report as clean.

manualHand-installed binaries
opt-appApps dropped in /opt
runtimeLanguage runtimes
What package managers miss

Discovery, not just enumeration

  • Python virtualenvs found by walking the filesystem for pyvenv.cfg, then reading each dist-info METADATA file.
  • Per-project node_modules trees found by locating package.json files outside node_modules.
  • Manually installed binaries that never touched apt, rpm, or an installer.
  • Applications dropped into /opt by a vendor script or a helpful contractor.
  • Language runtimes installed side-by-side with the distribution version.
Inventory — DEV-LNX-014
apt1,284 packages
/var/lib/dpkg
pip-venv96 packages
/srv/apps/billing/.venv
npm-pkg743 packages
/home/dev/portal/node_modules
opt-app1 unmanaged app
/opt/vendor-agent
manual4 unmanaged binaries
/usr/local/bin
5 unmanaged items — invisible to a package-manager-only scan.

Findings

Component, version, fix, and exactly where it lives

Each finding names the component, the installed version, the fixed version, the collector that found it, and how many devices carry it. No vague “host is vulnerable” — the row tells your engineer what to change.

Findings — Fleet (176 CVEs)
CVESeverityCVSSComponentInstalled → FixSourceDevices
CVE-2026-27820
CRITICAL9.8
zlib
RubyGems
3.1.13.1.2
gem3
CVE-2026-39892
Public
CRITICAL9.8
cryptography
PyPI
46.0.346.0.7
pip-venv2
CVE-2018-1285
CRITICAL9.8
log4net
NuGet
2.0.32.0.10
nuget1
CVE-2025-47273
HIGH8.8
setuptools
PyPI
49.1.378.1.1
pip6
CVE-2020-11023
KEV Public
MEDIUM6.9
jQuery
NuGet
1.5.13.5.0
nuget4
KEV — in CISA's Known Exploited Vulnerabilities catalog Public — component is reachable from the internet

Sorted by exploitability

KEV listing and internet exposure move a finding up the queue regardless of its CVSS score. A 6.9 that is being exploited on a public host beats a 9.8 on an air-gapped box.

Source-attributed

The Source column tells you which collector found it — gem, pip-venv, nuget, opt-app. That is the difference between a five-minute fix and an afternoon of hunting.

Fleet-aware

The Devices count shows blast radius immediately. One finding on six machines is one change rolled six times, not six tickets.

Supply chain

It doesn't just find vulnerabilities — it hunts for compromise.

A compromised npm or PyPI release is not a CVE. There is no advisory to match, no fixed version to upgrade to, and often no signal at all until someone publishes a post-mortem weeks later. By then the package has been installed on developer laptops, build agents, and production servers.

That is why conventional vulnerability tooling misses supply-chain attacks: it is built to compare version strings against a database of known-bad versions. If the database does not know yet, the scan comes back clean.

Security Validation looks for signs of compromise inside the language ecosystems themselves — npm, PyPI, RubyGems, and NuGet packages — across the whole fleet, every night. Not one build server. Not one repo. Every device that has the package on disk.

  • Sweeps npm, PyPI, RubyGems, and NuGet packages wherever they are installed.
  • Covers per-project virtualenvs and node_modules, not just global installs.
  • Runs nightly on every device — laptops included, not just CI.
  • Findings land in the same queue as CVEs, with the same device attribution.
Supply Chain — Nightly Sweep
Registries evaluatedFleet-wide
npmnpm · npm-pkg4,182
PyPIpip · pip-venv1,940
RubyGemsgem612
NuGetnuget388
Flagged for review
1 package release
Compromise indicators — no CVE assigned. Present on 2 devices.
Why nightly: a malicious release can be published, installed, and yanked inside a single day.
Remediation — Job Queue
65
Awaiting approval
8
Awaiting confirm
211 MB
Backups held
#87web-prod-04proposed
npm project
/var/www/portal
UNKNOWN1 pkg · 1 CVE
Dry run Approve
Packages (1)
follow-redirects1.15.9 1.16.0
MEDIUMCVE-2026-40895
Nothing runs unattended. Every job is proposed, reviewed, and approved by an admin — with a rollback backup held first.

Remediation

Remediation you approve, not remediation that surprises you

This can be automated. We generally recommend it is not. Upgrading a library inside a running application is a different risk from a routine OS patch, so by default Security Validation builds the remediation plan and proposes it rather than acting. Every job lands in a queue with the host, the exact path, the packages involved, the versions it wants to move between, and the CVEs it closes — then waits for an admin.

From there you have two moves. Dry run executes the plan without committing it, so you can see precisely what would change before anything does. Approve runs it for real. A backup is held first, so a patch that breaks a vendor app can be rolled back rather than argued about.

Updates come from the OS vendor or distribution repositories. The remediation path uses the same trusted sources as your normal patch process, so fixing a supply-chain problem never introduces one.

Closed-source software is handled honestly. If an application ships as a compiled or encrypted binary, nobody can patch its internals from the outside — the fix has to come from the vendor. Those findings are flagged as a manual upgrade so they land on a call with the vendor instead of silently failing in a queue.

And when an upgrade is genuinely awkward — a pinned dependency, a package with no clean fixed version — the Ask AI tab works the problem with you: what the finding affects, what is likely to break, and how to confirm the upgrade actually worked.

  • Every job is proposed first — nothing changes on a device without an approval.
  • Dry run any job to see exactly what would change before committing it.
  • Backups are held before a run, so a bad patch can be rolled back.
  • Jobs carry host, path, package versions, severity and CVE up front.
  • Re-plan rebuilds the queue against the latest scan when your infrastructure moves.
  • Updates pulled from vendor and distribution repositories only.
  • Closed-source binaries are flagged as a manual, vendor-side upgrade.
  • Ask AI helps you work through and validate the difficult upgrades.
  • Every attempt is recorded in the History tab.

Reporting

One click, full enterprise software inventory

Most organizations cannot answer “what software is installed here?” without a project. Because the nightly scan already collects it across every system, the answer is a button. Everything it finds — including software somebody installed by hand that no package manager ever registered — rolls into the Software Audit and Software Reporting automatically. And because the platform already manages those devices, you can push software out to any set of them from the same place.

Software Audit

Every install across every system lands in the audit — including manually installed applications that never went through apt, rpm, or an installer. That is the category most inventories quietly miss.

Software Reporting

Run a report over the whole fleet, dated to the scan, with versions and sources. CSV for your own analysis and asset reconciliation, PDF for the people who need a document.

Install, not just observe

Select devices and deploy software to them from the same console that found the gap. Reporting and remediation are the same system, so what you discovered is what you act on.

Nightly fleet scan
InventoryCVE correlationSupply-chain sweepCSV / PDF

The same data set drives the findings queue, the remediation runs, and the export. One collection, three uses.

Security Validation — Questions We Get

See what is actually installed on your fleet

Most first scans surface software nobody knew was there. Book a demo and we will walk the findings queue, the supply-chain sweep, and the export with your environment in mind.

← Back to AlertMonitor