redis-attack

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs retrieving and printing sensitive values from Redis (KEYS/GET/HGETALL/SCAN) and shows commands that include plaintext passwords (e.g., -a attempts and CONFIG SET requirepass "backdoor_pass"), so an agent following it would need to handle and output secret values verbatim.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). These URLs include multiple GitHub repositories that host offensive Redis exploit tools (module builders, rogue servers, SSH/webshell writers) plus a direct webshell URL (http://TARGET/shell.php), all of which are high-risk sources for distributing malware or enabling RCE.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This is an explicit offensive playbook for exploiting misconfigured Redis instances—providing step-by-step commands for RCE, persistent backdoors, credential theft, data exfiltration, and lateral movement.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs users to git clone and then run/compile code from https://github.com/n0b0dyCN/redis-rogue-server and https://github.com/n0b0dyCN/RedisModules-ExecuteCommand, which fetches remote code that is executed at runtime.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). This skill explicitly instructs writing webshells, cron jobs, SSH authorized_keys, loading malicious modules, and changing Redis configuration to persist backdoors and modify system files on targets, i.e., direct instructions to alter a machine's state and gain persistent access.