Network reachability analysis tells you whether an attacker can get to a vulnerable server over your network. Onitโs Reachability Agent reads your firewall, WAF, load balancer and cloud rules, matches them to the servers behind them, and returns a verdict with the path.
Your current scanner shows them as equals.
Severity ignores your network. TheReachabilityAgentreadsit.
When your network changes, it checks again.
How it works
How does network reachability analysis work?
InternetPartner linkAPP-SRV-14Reachable from the internetVPN-GW-02Restricted ยท 2 allowed IPsHR-DB-03Reachable internallyFIN-SRV-07Not reachableJust became reachableLAB-12UnknownA rule changed
Reads your network
Read-only: firewalls, WAFs, load balancers and cloud, as deployed.
Matches rules to assets
Links every rule to the assets behind it.
Traces from where attackers start
The internet, employee networks, partner links, and your own internal segments.
Returns a verdict with the path
A blind spot never passes as an all-clear. If config data is missing, the verdict is Unknown, never Not reachable.
Keeps watching
When a rule changes, it checks again and flags anything that just became reachable.
What does each reachability verdict mean?
Every answer, with its path.
VerdictWhat it meansWhat to do
Reachable from the internet
Open to any source
Fix now
Restricted ยท N sources
Reachable from the internet, but only from named sources
Short SLA, or confirm the allowlist
Reachable internally
Reachable from an employee network, partner link or another internal segment
This week
Not reachable
No path from any attacker starting point
Accept for now, with proof. Rechecked if anything changes.
Unknown
Not enough configuration data
Treat as reachable
Reachability applies to vulnerabilities exploited over the network. The agent recognizes client-side and local flaws, like a browser bug, and never closes them for lack of a path.
40 findings. 3 decisions.
A critical RCE lands on 40 servers. The agent splits them by path, and each group becomes one decision in the Decision Center.
6Reachable fromthe internetFix now12Reachable from theemployee networkThis week22In an isolated segment,no path inAccept for now, with proof.Rechecked if anything changes.6Reachable from the internetFix now12Reachable from theemployee networkThis week22In an isolated segment,no path inAccept for now, with proof.Rechecked if anything changes.
From a real environment ยท Oct 2026
Three scanners said nothing. One gateway was open to the internet.
Straight answers. Anything else, ask us in the demo.
Do you need access to my firewalls?
Read-only access to configurations. The Reachability Agent never writes a rule.
Do I have to install an agent?
No. Nothing is installed on your servers. Onit reads your firewall, WAF, load balancer and cloud configurations read-only, by API. To include internal firewalls that arenโt reachable from outside, you can add an optional Onit edge connector, a VM you deploy inside your network.
How current is it?
Full collection happens in minutes, and verdicts are recomputed whenever new configurations arrive.
What about attackers with stolen credentials?
This agent answers the network path. Identity paths are on our roadmap.
What is the difference between network reachability and code reachability?
Code reachability asks whether your application calls the vulnerable function. Network reachability, which this agent answers, asks whether an attacker can get to the vulnerable service over your network.
What happens to findings that arenโt reachable?
Unreachable vulnerabilities that need a network path to exploit close under a decision, with the path as evidence. They reopen if a rule change opens a path.
Find the criticals that have a way in.
See it in a 30-minute demo. Bring your network engineer.