VERISCO Alertio Respond
Response scripts
Write your own response code, check it for errors and dangerous commands, test it, then package it as a signed bundle and deploy it to your endpoints, all from the console.
When the built-in actions are not enough, write your own. A response script is the code an agent runs when a response action calls it. Scripts are written and checked on Response Scripts, then packaged and distributed from Script Deployment, both in the Respond group. Both require the Administrator role.
Writing a script#
New Script opens the editor. A script has:
| Field | Values |
|---|---|
| Name and filename | How it is listed, and the file installed on the endpoint. |
| OS family | Common, Linux, macOS or Windows. A common script is packaged for both Linux and macOS. |
| Language | Python, Bash, PowerShell or CMD. |
| Command | The command name a response action uses to call it. |
| Category and description | How it is grouped and explained in the library. |
The code is written in the editor pane. The inspector beside it has four tabs: Validation, Test, History and Details.
Validating and testing#
Validate checks the script and lists each finding with its location.
| Check | Looks for |
|---|---|
| Syntax | Errors that would stop the script from running. |
| Security | Dangerous constructs, such as deleting from the root, piping a download into a shell, or evaluating remote content. |
| Compatibility | That the script accepts its input the way the agent provides it, and handles both directions of an action. |
Two tests run the script against a sample alert: Test (add) runs the main path, and Test (delete) runs the rollback, so both directions of a reversible action are checked. Python and shell scripts are run in an isolated sandbox; PowerShell and CMD scripts are checked without being run.
The lifecycle#
A script moves through three states: Draft → Validated → Published. Publishing is blocked while validation fails, and only published scripts can be deployed. Every save keeps the previous version in History.
Deploying#
On Script Deployment, Build WPKs bundles every published script into signed installation packages. The control is unavailable until at least one script is published.
Choose who receives them:
| Scope | Targets |
|---|---|
| By OS group | Every endpoint running the chosen operating system. |
| Specific agents | Endpoints you select. Connected and offline agents are told apart before you commit. |
| Entire fleet | Every endpoint. Marked as the highest-impact option. |
Deployment asks for confirmation. Each target then reports its own state (pending, in progress, delivered, updated or failed), and the screen updates on its own as Alertio confirms the real state with each endpoint, rather than assuming success. Every rollout is kept in the history.
Once deployed, the script is available as a response action and its use is recorded like any other.
Signed packages
Packages are signed, and an endpoint installs only packages whose signature it trusts. If a deployment fails with a signature error, see Troubleshooting.
Something missing or wrong on this page? Tell us and we will fix it.