Version 0.1.4 is still preparing for the stores. Below is what the extension already does. Google Sheets and the AI assistant in this list are marked as after launch.
Clicking the Grabsavo icon in Chrome opens the panel with access to the page; the free version asks for fewer permissions
Templates keep collecting when a catalog grows; re-drawn pages are read again or flagged “May be incomplete”; a run that broke off can continue where it stopped
CSV for Russian Excel uses a decimal comma; file columns match the preview; cleaner column names
Restoring a backup keeps newer templates; “What’s new” after an update; clearer texts
v0.1.3
Coming to the stores
Free local beta: every local feature is unlocked — no license key, payment or trial; AI and Google Sheets are not available yet
Multi-page collection: a short last page is kept; a 404 mid-list, a tab taken elsewhere or a page that did not load ends as “May be incomplete”, not a clean finish
Slow pages are waited for, not collected twice; full-width digits are read as numbers; an interrupted run resumes without losing rows
Updates section in Settings: check for updates and “What’s new” after an update; data and settings are kept
Grabsavo collects no personal data: no account, no e-mail; collected data stays in your browser
v0.1.2
Coming to the stores
Free public beta: Excel, CSV, JSON, item pages, cleaning, templates and schedules without a key or expiring trial
Affirmative first-use terms acceptance is required before inspecting the page
AI and Google Sheets remain unavailable; optional future AI costs require a separate choice
Private/incognito windows are deliberately excluded so collected data cannot mix into the normal browser profile
A dedicated automated readiness gate now checks Chrome Web Store and Firefox AMO packages
Store copy, privacy disclosures and reviewer instructions now match the shipped local-first build
Store screenshots come from the real extension UI and the critical export path is keyboard-tested
v0.1.1
Coming to the stores
Numbered pagination now stops cleanly on the last page instead of turning a complete collection into an incomplete one
Scheduling and recovery from background-worker interruption are more reliable, with partial data preserved
Background diagnostics are available in Settings and no longer leave the panel waiting forever
Release builds are gated by Chromium E2E, stress tests, manifest permissions and package-size checks
v0.1.0
Coming to the stores
Auto-detected tables and lists, candidate switch, columns by clicking on the page
Five pagination kinds: “Next”, page numbers and URL templates, “Load more”, infinite scroll, scroll inside a box
Runs continue after a browser restart; pause and stop
Item pages: fields from schema.org, meta tags, spec tables and click-picking