redis-pentesting

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). These are personal/unofficial GitHub repositories that host Redis remote-exploitation tools (rogue master, module builder, SSH key injection) which are high-risk dual-use/malicious code capable of delivering RCE or malicious modules.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This guide explicitly documents step-by-step offensive techniques for unauthorized access, remote code execution, persistence, credential theft and data exfiltration — it is clearly malicious/offensive in intent.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs fetching and running remote exploit code (e.g., git clone https://github.com/n0b0dyCN/RedisModules-ExecuteCommand and references to https://github.com/n0b0dyCN/redis-rogue-server, https://github.com/Avinash-acid/Redis-Server-Exploit, https://github.com/captain-woof/redis-rce-ssh) which would download and execute external code at runtime, so these URLs are flagged.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly instructs writing files to system locations (webroot, /var/lib/redis/.ssh, cron spool), generating/placing SSH keys, loading malicious modules and changing Redis config to achieve persistent RCE and SSH access—actions that modify the host state and require/attempt to bypass privileged protections.