windows-kernel-exploits

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly instructs the agent to read, save, and report "new credentials or tokens found" and contains examples that embed plaintext credentials in commands (e.g., domain/user:Pass, adm1n/P@ssw0rd), which requires the LLM to handle and output secret values verbatim.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). All URLs point to exploit/PoC repositories and a raw PowerShell payload (vulnerable drivers, kernel exploits, DLL loaders and privilege-escalation tools) which are high-risk sources for malware distribution and should not be downloaded or executed on untrusted or production systems.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document is an explicit offensive playbook for exploiting Windows kernel vulnerabilities and privileged operations — it provides step-by-step instructions for local and remote RCE, token theft, loading vulnerable drivers (BYOVD), EDR/PPL bypass, named-pipe impersonation, credential-stealing (Mimikatz), and persistence techniques, clearly enabling deliberate malicious abuse.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill includes explicit runtime commands that fetch and execute remote code—for example "git clone https://github.com/helviojunior/MS17-010" followed by running its scripts, and an inline PowerShell download+execute using "https://raw.githubusercontent.com/EmpireProject/Empire/master/data/module_source/privesc/Invoke-MS16032.ps1"—so these URLs are used at runtime to execute remote code.

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

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The prompt explicitly instructs loading vulnerable kernel drivers (sc create/start), copying/modifying system DLLs in System32, creating admin users, and performing kernel/local privilege escalation techniques that directly modify system state and require/seek elevated privileges.