gpt-pp-team-protocol-replay
Fail
Audited by Socket on Apr 29, 2026
1 alert found:
MalwareMalwareSKILL.md
HIGHMalwareHIGH
SKILL.md
MALICIOUS. The skill’s actual footprint is fundamentally incompatible with a normal agent skill: it automates payment/account creation flows, bypasses hCaptcha and anti-fraud controls, acquires refresh tokens, and supports continuous account-pool maintenance. Even with some official endpoints and partially legitimate installs, the overall capability set is purpose-built for abuse and credential/token handling beyond any proportionate security-research helper.
Confidence: 95%Severity: 97%
Audit Metadata