Summary

Implement automation for exporting and importing Chrome bookmarks.

The goal is to provide a simple shell-based solution that allows bookmarks to be backed up into the repository and restored later with minimal user interaction.

This feature is intended for personal backup and synchronization workflows.

Goals

• Implement using POSIX shell (.sh)
• Avoid Python unless absolutely necessary
• Support macOS as the primary platform
• Keep implementation simple and transparent
• Store exported bookmarks inside the repository
• Allow restoring bookmarks from a previously exported file

Folder structure

scripts/

    chrome/

        export-bookmarks.sh

        import-bookmarks.sh

        wait-download.sh

data/

    bookmarks/

        bookmarks.html

Export workflow

close chrome (optional)

↓

open bookmark manager

↓

navigate to

Bookmarks Manager

↓

open overflow menu

↓

Export Bookmarks

↓

wait for download

↓

move downloaded html

↓

data/bookmarks/bookmarks.html

Import workflow

verify bookmarks.html exists

↓

open bookmark manager

↓

open overflow menu

↓

Import Bookmarks

↓

select bookmarks.html

↓

wait until completed

↓

done

Assumptions

• User is already logged into Chrome
• User uses the default Downloads directory
• User has permission to automate Chrome
• User can grant Accessibility permissions if required
• Existing bookmarks may be merged by Chrome

The script does not attempt to manage synchronization conflicts.

Script responsibilities

export-bookmarks.sh

• Open Chrome if necessary
• Open the bookmark manager page
• Trigger bookmark export
• Wait until an HTML file appears in Downloads
• Move the exported file to:

data/bookmarks/bookmarks.html

• Replace the previous backup if one exists

import-bookmarks.sh

• Verify backup exists

data/bookmarks/bookmarks.html

• Open Chrome bookmark manager
• Trigger bookmark import
• Select the backup file
• Wait for completion
• Exit successfully

wait-download.sh

Utility script responsible for polling the Downloads directory.

Pseudo workflow:

loop

    if newest html file exists

        return path

    sleep 1

timeout after configurable period

Future improvements

• Timestamped bookmark backups

bookmarks/

    2026-06-10.html

    2026-07-01.html

    latest.html

• Compression into archive

bookmarks.zip

• Automatic diff between backups

• Restore specific snapshot

• Multiple Chrome profiles

Default

Work

Personal

• Chromium support

• Brave support

• Arc support

CLI

./scripts/chrome/export-bookmarks.sh

./scripts/chrome/import-bookmarks.sh

Later these can be wrapped by:

shuttle bookmarks export

shuttle bookmarks import

Definition of done

• Export script creates data/bookmarks/bookmarks.html
• Import script restores from data/bookmarks/bookmarks.html
• Scripts are idempotent
• Existing backup is overwritten safely
• Errors are handled gracefully
• Timeouts prevent infinite waiting
• Scripts are documented and runnable from a fresh clone