Skip to content

Behind the scenes — the Action Journal

Opened from the footer's graduation-cap button (or the command palette), the journal is the app's own behaviour as the subject — not the repository's contents, so it needs no git data of its own: every command was already recorded as it ran, and this reads that log back. Every row shows the real command line with no model configured at all; asking "Explain" is what turns it into a lesson, and the copy icon beside it takes the raw commands as-is, for pasting into a terminal or a bug report.

Explaining the commands behind a recent action

Every action you perform — a commit, a checkout, a rebase step — is recorded as the git command(s) it actually ran, newest first. Picking one and asking to explain it reads those commands back as a short lesson in plain language, remembered afterwards so reopening the same action doesn't regenerate it.

Explaining the commands behind a recent action

Captured from the running app by this scenario.

Do this

  1. Select the working-tree changes in the graph
  2. Enter the commit message test: action journal e2e marker
  3. Commit the staged changes
  4. Open the action journal
  5. Filter the action journal for action journal e2e marker
  6. Open the filtered action
  7. Click the copy-commands button
  8. Click the explain-action button

You should see

  • The repository HEAD commit subject is test: action journal e2e marker
  • The copy-commands button confirms the copy
  • The action explanation shows a finished explanation

This page is generated from apps/e2e/features/action-journal.feature, the end-to-end test that drives the feature it describes.

MIT licensed · Screenshots exported from the real app by the end-to-end suite.