The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/agent.py executes system-level administrative commands using the subprocess module, including iptables for network control and powershell for Active Directory management.\n- [DYNAMIC_EXECUTION]: In scripts/agent.py, Active Directory management functions dynamically construct PowerShell strings using f-string interpolation of arguments such as username. This pattern is susceptible to command injection if inputs contain malicious payloads designed to break out of the command string context during evaluation by the PowerShell interpreter.\n- [PRIVILEGE_ESCALATION]: The skill implements high-privilege operations, including the modification of network routing via iptables, altering the system /etc/hosts file, and disabling or resetting passwords for domain-level accounts. These actions have a high potential for impact if misused or triggered against unauthorized targets.\n- [INDIRECT_PROMPT_INJECTION]:\n
Ingestion points: The skill is designed to process untrusted telemetry data, such as IP addresses, hostnames, and account names, which are used as parameters for containment actions.\n
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the scripts or documentation.\n
Capability inventory: The agent has the ability to isolate hosts from the network, disable user accounts, and modify system configuration files.\n
Sanitization: The script lacks validation or escaping for the parameters interpolated into administrative command strings.

containing-active-breach