Security

Boundaries
& Verification

Security isn’t a slogan. It’s what the system refuses to expose, what it checks, and what happens when something fails.

Least exposure

The public surface is intentionally small.

Internal services stay private. Only the entry points that genuinely need to be public are public — everything else sits behind that boundary, unreachable directly from the outside.

Restrict administrative surfaces

Administrative access is network-restricted, not just password-protected.

A correct password from the wrong place still doesn’t get in.

General public request
denied
Admin boundary
restricted
Rate limit at the edge

Abuse gets stopped before it reaches application code.

Public API paths enforce request-rate limits at the edge, so a burst of runaway or abusive traffic doesn’t reach the systems doing real work.

Harden & check, on a schedule

Posture gets checked on a recurring schedule, not once at launch.

Host hardening checks

Configuration posture is checked on a recurring schedule, not just at launch.

Intrusion / rootkit scanning

Scheduled scans look for signs of unauthorized change.

Dependency & system checks

What the system depends on is checked, not assumed safe forever.

Verify before trust

Security posture is something to keep proving, not something declared once.

That happens through health checks, scheduled scans, operational logs, and ongoing monitoring — not a certificate hung on the wall.

Health checks
Scheduled scans
Operational logs
Ongoing monitoring
Recoverable by design

When something fails, there’s a way back and a record of what happened.

Backups, rollback paths, and operational evidence are shared infrastructure across every system here — not reinvented per product.

View Infrastructure →
Evidence boundaries

Claims here are limited to controls that can actually be verified.

  • No invented compliance or certification claims.
  • No "zero trust" language without something specific behind it.
  • No penetration-test claim without an actual test to point to.
  • Content Security Policy is currently in report-only mode while it is rolled out — not yet claimed as fully enforced.

Need a system that treats security as ongoing work, not a checkbox?

Explore RailTaskLite