The vulnerability stems from two key phases: 1) Storing malicious config via page edit (handled by SaveResponder), and 2) Executing commands using these configs (handled by TestResponder/TestRunner). The exploit demonstrates direct control over COMMAND_PATTERN/TEST_RUNNER through pageContent parameter in saveData requests. Functions involved in processing these untrusted variables during save operations (SaveResponder.saveContent) and command construction (TestResponder.makeCommand, TestRunner.getCommand) would appear in runtime profiles when malicious payloads are processed. Confidence levels reflect the exploit's clear demonstration of SaveResponder involvement and typical command injection patterns in Java web apps.