pass-the-hash

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly requires inserting NTLM hashes, AES keys, and ticket contents verbatim into commands and code (e.g., -aesKey 2ef70... and -hashes :NTHASH), so the LLM would need to handle and emit secret values directly.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The document is explicit malicious operational guidance: it teaches credential theft and abuse (NTLM hashes, AES keys, Kerberos tickets), covert lateral movement techniques (Pass-the-Key/Pass-the-Hash/Pass-the-Ticket), remote code execution (PsExec, WMI, services, scheduled tasks), and data/loot transfer (evil-winrm uploads/downloads), which together enable unauthorized access, persistence, and exfiltration.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I flagged only true high-entropy, literal keys that could be used directly for authentication. Specifically:
• Flagged:
• 2ef70e1ff0d18df08df04f272df3f9f93b707e89bdefb95039cddbadb7c6c574 — 64 hex chars (32 bytes), used as an AES256 key in getTGT.py and Rubeus examples. This is a high-entropy, usable credential.
• bc09f84dcb4eabccb981a9f265035a72 — 32 hex chars (16 bytes), used as an AES128 key in a Rubeus example. This is a high-entropy, usable credential.
• Ignored (not flagged) and why:
• NTHASH — documentation placeholder (explicit placeholder token).
• aad3b435b51404eeaad3b435b51404ee — known LM-empty constant used in examples (standard placeholder), not a secret to flag.
• 'Password123' and similar simple passwords shown in examples — low-entropy example/setup values, therefore ignored.
• Other tokens like ':NTHASH', 'YOUR_*' style placeholders and truncated/redacted forms are documentation placeholders per the guidance.

Given the presence of the two literal AES keys above (usable to request TGTs), this documentation contains active high-entropy secrets.

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 obtaining sudo/root (e.g., sudo ntpdate), modifying local system files (/etc/hosts), and running actions that create services or inject credentials (which alter system state), so it pushes the agent to perform privileged state-changing operations.