The gpt-terminal-cli package (version 1.0.0) masquerades as an AI chat CLI but executes a post-install script that launches a detached implant daemon. This implant maintains persistence by respawning up to 50 times if terminated and communicates with a hardcoded C2 server at http://13.60.13.215:7771 using an AES-256-GCM encrypted and HMAC-signed protocol. It supports multiple malicious operations including reverse PTY shell access, credential theft, clipboard and keylogging data collection, file exfiltration, lateral movement, privilege escalation, and anti-forensics. The implant dynamically updates its C2 endpoint and encryption keys by polling DNS TXT records every 30 minutes, enabling infrastructure agility without package updates. On Windows, it bypasses AMSI protections by injecting a reflection stub into PowerShell invocations.

This malicious package enables attackers to gain persistent remote access to infected systems, steal sensitive credentials and data, monitor user activity via keylogging and clipboard capture, exfiltrate files, escalate privileges, move laterally within networks, and evade detection through anti-forensics techniques. The implant's self-healing and dynamic C2 rotation capabilities increase its resilience and complicate incident response.

No official patch or remediation is currently available. Users should immediately uninstall version 1.0.0 of gpt-terminal-cli and avoid installing this package. Conduct thorough system scans for the implant and related artifacts if this package was installed. Monitor network traffic for connections to the known C2 IP (13.60.13.215) and related domains. Since this is a malicious package, remediation involves removal and incident response rather than patching. Check vendor or repository advisories for updates or warnings.