Script workspace

Scope and prerequisites

The unified script workspace is announced in the v3.2.17 release notes for its macOS and Windows packages. This guide uses that release as the earliest confirmed publication evidence, not the earliest implementation date. Sending requires an online SSH target.

Create a reusable script

1. Choose Tools → Script Library. The left side contains library navigation, categories and scripts; the main area shows script content and targets, with history and output below.
2. Create a script named “Directory check”. Use pwd as the body and keep batch sending disabled for this first example.
3. Select a category, review the risk and confirmation settings, then save. The language setting controls highlighting; it does not choose a remote interpreter.
4. Search or select the saved script and review its saved content.

Send and inspect

1. Choose one connected test SSH session. Confirm its host and split identity.
2. Send and review any confirmation. Do not edit or send an unsaved draft as if it were the saved script.
3. Select the new history record. Inspect captured output and its truncation or capture information.
4. Check the actual remote terminal. “Sent” is a transport status, not proof that the command or business task succeeded. Manual result marking is separate.
5. If batch sending is later enabled, select each target explicitly. On partial failure, inspect and retry only failed targets to avoid duplicates.

Common problems

No online targets means SSH must be connected first. A disabled script must be enabled before sending. If saving fails, keep the draft and fix storage permissions or space. Deleting a script does not undo a command already sent, and historical output may be truncated. AI generation is a separate service-dependent capability.