Insights / STIGs
STIGs

STIG'd: hardening as a pipeline.

What Security Technical Implementation Guides are, why manual compliance decays, and how automation turns hardening into a system property.

Ausper Technology · July 20, 2026 · Insight · STIGs · Hardening

A STIG, Security Technical Implementation Guide, is DISA's hardening checklist for a specific technology: hundreds of configuration rules for your OS, database, browser, or network device, each with a severity (CAT I/II/III) and a check-and-fix procedure. 'Get STIG'd' is the standing order of DoD system administration.

Why they hurt

Volume and drift. A single RHEL server carries hundreds of rules; an environment carries thousands of rule-instances. Manual hardening takes days per system and decays immediately; every patch and every change re-opens findings. Programs that treat STIGs as a quarterly manual event live in permanent catch-up.

The way out

Automate application: Ansible or comparable tooling applies baselines identically, every time. Automate verification: SCAP scans on schedule, results flowing into your POA&M and dashboards automatically. Handle deviations formally: some rules legitimately can't apply, so document the deviation, the mitigation, and the approval once, instead of re-arguing it every assessment. Hardening becomes a pipeline property, which is what continuous ATO requires.

Common questions

What is a STIG?
DISA’s configuration standard for a specific product, published as a checklist of requirements rated CAT I, II, or III by severity. Tooling reports each one as open, not applicable, or not a finding.
Why does manual STIG compliance decay?
A hand-hardened host is only correct on the day it was hardened. Patches reset settings, installs change them, and ordinary administration drifts them, and nothing notices until the next scan.
How do you automate hardening?
Put the configuration in the image rather than in a runbook, enforce it with configuration management, and scan inside the pipeline so a non-compliant build never reaches an environment. A drifted host then becomes an instance to replace, not an incident to investigate.

Related reading

Put this to work

Working on this now?

Tell us where your program stands and we'll tell you what we would do next.

Talk to Ausper