Auto_JB_APE

jb_ape is an automated red-team engine designed to perform security assessments by probing target defenses, generating and mutating attack payloads through browser or API interfaces. It features a unique three-tier judgment system that ensures machine-verified outcomes for every attempt, employs a controlled submission budget to enhance efficiency, and utilizes reinforcement learning techniques to optimize the attack strategy while avoiding guessing. This tool is intended strictly for authorized use in sanctioned environments such as penetration testing or capture-the-flag competitions.