Automated Response Overview

Automated response (SOAR, short for Security Orchestration, Automation and Response) lets UTMStack v12 act on an alert within seconds. You build flows: when a new alert matches a flow's conditions, the flow's steps run, for example a command on the affected machine through its UTMStack agent, a call to another system, an email, an in-app notification or a new incident.

What a flow is

A flow has a name, a description, trigger conditions, and a set of steps linked by wires. You draw it on a canvas in SOAR > Flows.

ConceptWhat it means
FlowOne automated response: trigger conditions plus the steps to run.
Trigger conditionsThe alert fields a new alert must match. A flow runs only when all of its conditions match.
Step (the editor calls it a node)One action: run a shell command, call a web address, ask the AI, send a notification, open an incident, send an email, or check a condition.
Executor stepA step that does something (for example runs a command).
Enrichment stepA step that fetches data that later steps can use (an HTTP call or an AI answer).
Success wire / error wireEach step has two exits. The green wire runs when the step succeeds, the red wire when it fails.
Start stepsThe steps wired directly from the trigger. They all start at the same time.
RunOne execution of a flow for one alert. Every step in a run gets its own row in Execution history.
AgentThe UTMStack agent on a machine. Shell steps run through it.
PlaceholderA value from the alert, written $(alert.<field>), filled in when the step runs.
VariableA stored value, written $[variables.NAME], for example a token or a host name. Variables can be secret.

How a flow runs

flowchart LR
    A[Correlation rule raises an alert] --> B{Matches all conditions of a flow that is switched on?}
    B -->|No| H[Nothing happens]
    B -->|Yes| C[UTMStack creates a run]
    C --> D[Start steps run]
    D -->|Success| E[Steps on the green wires]
    D -->|Failure| F[Steps on the red wires]
  1. A correlation rule raises a new alert.

  2. The SOAR engine plugin checks the alert against every flow that the alert's tenant has switched on. Repeat alerts (echoes, which UTMStack groups under an earlier alert) never start a flow.

  3. If all conditions of a flow match, the server creates a run and starts every start step.

  4. When a step finishes, UTMStack follows its green wires if it succeeded, or its red wires if it failed. The branch that is not taken is recorded as Unreachable.

  5. Each step and its status appear in SOAR > Execution history.

The server picks up new work as soon as it arrives and also checks for waiting steps every 15 seconds. It runs up to 5 steps at the same time, and one step may take up to 5 minutes. In testing of UTMStack v12.0.1, a command reached the machine about a quarter of a second after the alert.

Where to find it

The SOAR menu has three pages:

PageWhat you do there
FlowsList, create, edit, copy, switch on and off, and start flows by hand. See Flows Page and Building a Flow.
Execution historySee every step that ran, its status and its target machine. See Execution History and Run Statuses.
Interactive ConsoleRun commands live on one agent. See Interactive Console.

The old addresses /soar/console and /soar/audit still work: the first opens the Interactive Console, the second opens Configure > Settings > App Audit Logs.

Who can do what

TaskPermissionBuilt-in roles that have it
View flows, execution history and variablessoar.readAdministrator, Analyst, Viewer
Create, edit, delete, switch flows on and off, manage variables, use the Interactive Consolesoar.writeAdministrator, Analyst
Start a flow by hand (Trigger flow, the alert response button)soar.executeAdministrator only

Automated response does not need an Enterprise licence.

Tenants

On a multi-tenant installation each tenant has its own flows, variables and history, and never sees another tenant's flows. The flows that ship with the product are shared, but each tenant switches them on separately. The instance operator (an Administrator of the default tenant) can also push a flow to many tenants at once with Apply to all tenants.

What changed from UTMStack v11

  • Workflows are now called flows and are drawn as a graph of steps, not a list of shell commands joined with ;, && and ||.

  • Steps are no longer only shell commands. There are HTTP call, AI enrichment, notification, incident, email and conditional steps.

  • Placeholders start with alert.: $(adversary.ip) is now $(alert.adversary.ip), and $(log.message) is now $(alert.lastEvent.log.message).

  • The catalogue of 73 action templates and the 23 playbooks are replaced by 94 shipped flows. Most of them have no trigger and are meant to be copied. See Shipped Flows.

  • The Automation Variables page is gone. Manage variables from the Variables button in the shell step editor or in the Interactive Console.

  • Statuses changed: see Execution History and Run Statuses.

Before you switch a flow on

Flows act for real. They block addresses, disable accounts, stop processes, isolate machines and shut them down. Test every flow by hand with Trigger flow on a test machine before you switch it on.

  • Every matching alert starts a new run. There is no limit and no grouping of runs, so a noisy rule can start the same flow many times.

  • A flow you create from scratch is saved switched on and starts running at once. A copy of an existing flow is saved switched off. See Flows Page.

  • Shell steps need the agent to accept remote commands. An agent installed with remote control disabled refuses every command, and this can only be changed by reinstalling the agent on that machine.

  • To tell a failed command from a successful one, the agent must be from UTMStack v12.0.2 or later. Older agents report a failing command as executed, so the error wire never runs.

  • Flows Page

  • Building a Flow

  • Step Types

  • Placeholders

  • Step Outputs and Variables

  • Shipped Flows

  • Running Steps on Another Host

  • Interactive Console

  • Execution History and Run Statuses

  • Use Cases

  • Worked Examples: Responding Through Other Systems

  • Troubleshooting Flows