The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains numerous commands for executing penetration testing tools such as netexec, nmap, responder, ntlmrelayx.py, and certipy. These commands are used for network scanning, hash capturing, and triggering authentication on remote targets.- [PRIVILEGE_ESCALATION]: Several instructions require the use of sudo to run network tools like responder. Additionally, the skill instructs the user to modify system-level configuration files such as /etc/responder/Responder.conf and /etc/proxychains4.conf to facilitate attacks.- [DATA_EXFILTRATION]: The methodology focuses on capturing NetNTLM hashes and dumping sensitive credential data (SAM, LSA, DPAPI) from remote machines using tools like secretsdump and lsassy.- [EXTERNAL_DOWNLOADS]: The skill references multiple external scripts and tools (e.g., PetitPotam.py, dementor.py, dfscoerce.py) that are not part of a standard distribution and would need to be acquired from third-party sources.- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process data from external network services and tool outputs (e.g., LDAP/SMB scan results). If these external sources are controlled by an adversary, they could potentially influence the agent's behavior via data-driven instructions, although this is a inherent risk of the task type.

ntlm-relay-attack