> ## Documentation Index > Fetch the complete documentation index at: https://docs.onefirewall.com/llms.txt > Use this file to discover all available pages before exploring further. # Clean vs Malicious > Of the traffic a firewall already allowed through, how much OneFirewall's intelligence identifies as malicious Of the **16,055** events the client's firewall allowed through, OneFirewall's scoring identifies **14,116** as clean and **1,939** as malicious. *** ## Scoring after the fact, not instead of This isn't a re-evaluation of a blocking decision. It's a second, independent classification applied to traffic that already passed. The firewall made an allow/deny call based on its own rules; OneFirewall separately matches the same traffic against the Alliance's threat intelligence and Crime Score data. The two systems can disagree, and this chart is where that disagreement becomes visible: **1,939** events, roughly 12% of everything permitted, carried an indicator the existing rule set had no way to recognize. ## Where this number goes next **1,939** isn't a flat category. It's the starting point for the severity breakdown that follows: Low, Medium, High, and Critical. That breakdown is what turns "this was malicious" into a prioritized list of what to act on first. *** Proof of Value engagements run this same clean-versus-malicious split against a client's own permitted traffic. [Start a Proof of Value](https://onefirewall.com/proof-of-value).