LinearAgent
Bases: BaseAgent
This agent will perform a list of actions, and either repeat them indefinately or Sleep. If no action list is given it will always Sleep.
This agent is intended to help with stress testing and debugging, as it allows the tester to perform a series of specific actions that they have pre-listed. As actions sometimes require other action to have been performed successfully to function, this sequential execution makes it possible to test these actions. The agent can also be used to test that another type of agent is reacting correctly to a specif series of actions.
Attributes:
| Name | Type | Description |
|---|---|---|
action_list |
Dict[int:Action, dict]
|
a list of actions, with parameters, to be executed sequentially. |
circular |
bool
|
if true, the list of actions will be repeated once the end is reached, otherwise the remaining actions will be Sleep |
step |
int
|
the agent's internal step counter (not accurate to the State) |
last_action |
str
|
the name of the previous action that was performed |
print_action_output |
bool
|
print the action and action success |
print_obs_output |
bool
|
print the observation from the action |
Example
An action list could look like the following:
action_list = {
0: PrivilegeEscalate(
hostname='contractor_network_subnet_user_host_2',
session=0,
agent='red_agent_0'),
1: DiscoverRemoteSystems(
subnet=IPv4Network('10.0.0.1/24'),
session=0,
agent='red_agent_0'),
2: AggressiveServiceDiscovery(
session=0,
agent='red_agent_0',
ip_address=IPv4Address('10.0.0.2')
}