See what one Linux host still needs before you call it fixed.
Run a read-only Patch Truth Snapshot and get a short report that says whether the host needs package updates, a restart or reboot, nothing, or a better scan.
curl -fsSL https://get.executepath.dev/snapshot.sh | sudo sh
Installed EVRs are below the vendor-fixed package versions.
"Updated" and "fixed" are not always the same thing.
Sometimes the package really is behind. Sometimes the package is updated but a service still needs a restart. Sometimes the host is clean. The report should say which one happened instead of handing you another generic CVE pile.
fixed EVR / 3.0.7-27.el9
on disk -> fixed / running -> restart required
A quick scan, not a deployment project.
Run one command
A read-only snapshot collects the minimum package, OS, kernel, and runtime evidence needed for a host-level package-truth report.
Get the report
oxharden analyzes the snapshot, writes a shareable report URL, and shows the exact evidence behind the result.
Decide whether to expand
Use one host as the fast proof point, then move to the full fleet agent only when the signal is worth tracking continuously.
The report has four honest outcomes.
The goal is not to dramatize every host. The goal is to give an operator a concrete next step, including when the data is not good enough to trust.
Package update required
The installed package is below the vendor-fixed EVR. This is the straightforward patch-now case.
Restart or reboot required
The package may be fixed on disk, but runtime evidence still points to old code, deleted libraries, or kernel debt.
Clean snapshot
No vulnerable or restart-required package findings were found in the point-in-time scan.
Incomplete or unsupported
The report refuses to pretend uncertainty is clean: unsupported distro, partial collection, or evidence gaps are called out.
A report your team can actually use.
The output is designed for the messy handoff: ops sees commands and package evidence, security sees CVE context, and a manager gets a short summary without asking for a dashboard tour.
Installed EVRs are below the vendor-fixed package versions.
Start with one host. Expand only when the signal is real.
Snapshot mode is the low-friction proof. Fleet mode is the continuous product for package truth, exposure, compliance, and trend history across many hosts.
Built for cautious production evaluation.
Get a package-truth report before the next patch meeting.
Run the one-command snapshot on a representative RHEL-family box and see whether the next step is update, restart, clean, or rerun.
curl -fsSL https://get.executepath.dev/snapshot.sh | sudo sh
Questions, answered.
It runs a one-time read-only collection on a single host, uploads the package and runtime evidence for analysis, generates a report, and exits.
No. Snapshot mode is the first-look diagnostic. Fleet mode is the persistent agent and dashboard for continuous monitoring across many hosts.
Package versions are necessary, but they do not always prove remediation is live. Some fixes require service restarts or host reboots before old code stops running.
The report calls that out as incomplete or unsupported instead of giving a false-clean result.