jazzer

The skill is internally consistent with JVM fuzzing and shows no evidence of credential theft or malicious data exfiltration. Its primary risk is enabling an AI agent to perform security testing against selected code; use it only on authorized targets and isolated workloads.