History Search & Backup

BrowseVault is a local-first Chrome extension project for private browser history search, backup, export, and preservation.

The working Chrome Web Store title is:

BrowseVault: History Search & Backup

The short positioning is:

Search, back up, export, and preserve your browser history locally.

This repository contains a working Manifest V3 extension implementation. It includes:

• a loadable Chrome extension shell;
• a full-page BrowseVault app;
• a tabbed interface that keeps History results first and moves secondary tools out of the scroll path;
• local IndexedDB archive storage;
• background capture for new Chrome history visits;
• manual sync from currently available Chrome history, expanded to individual visits where Chrome exposes them;
• atomic Chrome sync and live-capture metadata commits so recorder health matches recorded rows;
• visible archive health for startup, Chrome sync, live capture status, malformed vault rows, tombstones, and duplicate active records;
• query syntax for site/host/domain, title, URL, manual category/tag, source, transition, visit count, exact-day/date ranges, local-hour filters, exclusions, wildcards, phrases, and regex;
• bounded typo-tolerant matching for longer plain keyword searches;
• cancellable worker-backed, chunked local search scanning for large vaults, covered by synthetic large-history tests;
• bounded result retention for limited searches, covered by a 100k-record synthetic history test;
• one-click Today, Yesterday, 7 Days, 30 Days, and All Dates shortcuts that fill ISO-style date filters;
• saved searches for repeated local history queries, date/limit filter sets, and newest/oldest result ordering;
• highlighted matches in vault and Quick Open search results;
• keyboard-first search with autofocus, global query focus, Enter search, and debounced live refresh;
• keyboard navigation for visible history results with arrow keys, Enter-to-open, Space-to-select, and Shift+Space range selection;
• local-date grouped vault results for easier timeline scanning;
• Quick Open search across open tabs, bookmarks, downloads, recently closed tabs, and closed windows, with source-aware switch/open/open-in-background/restore actions, keyboard navigation, and readable source-unavailable warnings;
• toolbar and keyboard-command opening that creates a new BrowseVault tab unless the current active tab is already BrowseVault;
• cross-tab vault refresh for deletes, imports, cleanup, reset, rules, manual sync, live Chrome-history capture, and native history removals, using BroadcastChannel plus a Chrome local-storage revision fallback;
• Settings escape hatch for opening Chrome's native history page without making BrowseVault a default-history override;
• JSON backup export and import, with chunked JSON serialization and integrity payload generation for large backup downloads;
• deterministic newest-first full exports, plus current-result exports that follow the selected newest/oldest result order, including chunked formula-safe CSV generation with ISO timestamp, local date/time, manual category, visit id, and Chrome id fields;
• chunked offline HTML exports with summary metrics, exact timestamps, manual category column, in-file filtering, sortable columns, and safer link handling;
• visible backup status with freshness, format, record count, file size, restore confidence, checksum details, and a backup self-test that verifies restorable rows;
• visible storage self-check status for the local vault metadata layer;
• secondary archive insights for top domains, busiest day, active days, and date range;
• recent activity log for completed backup, export, import, cleanup, delete, restore, rule, and reset operations;
• SHA-256 integrity metadata for JSON backups;
• staged import preview with valid row, duplicate row, existing visit, new visit, rule, and restore-check health status;
• atomic import commits that write restored visits, imported rules, and import metadata together;
• imports preserve existing local vault tombstones and Chrome-deletion markers when matching visits are imported again;
• imports merge duplicate archive rows into unique restored records and report duplicate rows in the final status/activity log;
• JSON, CSV, TSV, Google Takeout Browser History, Google My Activity, and common competitor-export import handling;
• current filtered result-set JSON, CSV, and HTML export without manual selection;
• current filtered result-set vault deletion with confirmation and undo;
• current filtered result-set Chrome history deletion by URL with confirmation;
• selected-record JSON, CSV, and HTML export;
• selected-record URL copy;
• selected-record batch opening with a tab-flood safety cap;
• bulk deletion from the BrowseVault vault;
• selected-domain blacklisting for future archive capture;
• checkbox, Shift-click, and Shift+Space range selection, plus select visible, invert visible, or select all filtered vault results;
• sticky result controls with incremental Load More, one-click Show All, and Top/Bottom jumps for large result sets;
• undo for the last BrowseVault vault deletion;
• optional URL-level deletion from Chrome history for selected records or the current filtered result set;
• visible Reset Vault action for full local BrowseVault data reset without touching Chrome history;
• domain blacklist, whitelist, and manual category rules;
• manual retention cleanup that previews old vault records and keeps whitelisted domains;
• duplicate cleanup that previews repeated active vault records and moves extras to undoable deletion;
• local preferences for system/light/dark theme, accent color, high contrast, text size, date format, default result limit, backup reminders, backup/export save mode, backup filename prefix, and backup filename template;
• backup reminder status follows the configured reminder interval and warns when the next reminder date is reached;
• exact ISO visit timestamps in history result metadata while visible rows follow the selected date format;
• in-app trust, permission, and product-limit disclosures;
• local-first privacy documentation;
• automated folder-density checks that enforce 12-file runtime/support/docs and 15-file feature-folder budgets;
• versioned release notes in CHANGELOG.md, including trust-sensitive behavior changes;
• generated PNG extension icons;
• generated Chrome Web Store promo PNGs;
• deterministic runtime-only local ZIP packaging that excludes repo research, tests, scripts, StorePilot files, and feature docs;
• copied research and product docs in docs/;
• Chrome Web Store listing notes in store/listing.md;
• StorePilot-ready Chrome Web Store import files in store-listing/chrome-web-store/;
• repository description and topic/tag metadata in package.json and docs/project/repository-metadata.md.

.
├── AGENTS.md
├── _locales/
│ └── en/
│ └── messages.json
├── docs/
│ ├── architecture/
│ ├── project/
│ ├── release/
│ ├── research/
│ ├── chrome-web-store-additional-fields.md
│ ├── chrome-web-store-category.md
│ ├── chrome-web-store-media.md
│ ├── chrome-web-store-privacy-form.md
│ ├── README.md
│ └── storepilot-project-structure.md
├── assets/
│ └── icons/
├── scripts/
│ ├── check-file-sizes.mjs
│ ├── check-folder-density.mjs
│ ├── check-imports.mjs
│ ├── check-locales.mjs
│ ├── check-manifest-paths.mjs
│ ├── check-privacy-permissions.mjs
│ ├── check-syntax.mjs
│ ├── generate-icons.mjs
│ ├── media/
│ ├── playbook/
│ ├── qa/
│ ├── package-extension.mjs
│ ├── validate-extension.mjs
│ ├── verify-package.mjs
│ └── zip-utils.mjs
├── src/
│ ├── app.css
│ ├── app.html
│ ├── app.js
│ ├── background.js
│ ├── browser-memory.js
│ ├── export-format.js
│ ├── features/
│ │ ├── activity-log/
│ │ ├── app-shell/
│ │ ├── background-runtime/
│ │ ├── backup-import/
│ │ ├── browser-memory/
│ │ ├── display-preferences/
│ │ ├── history-export/
│ │ ├── history-results/
│ │ └── vault-management/
│ ├── platform/
│ ├── query.js
│ ├── styles/
│ └── storage.js
├── store/
│ └── listing.md
├── store-listing/
│ └── chrome-web-store/
│ ├── listing/
│ │ └── en.md
│ └── media/
│ ├── icon-128.png
│ ├── promo/
│ │ ├── marquee-promo.png
│ │ └── small-promo.png
│ └── screenshots/
├── test/
│ ├── features/
│ ├── platform/
│ ├── query/
│ ├── scripts/
│ └── storage/
├── manifest.json
├── package.json
├── PRIVACY.md
└── README.md

1. Open chrome://extensions.
2. Enable Developer mode.
3. Click Load unpacked.
4. Select this repository folder.
5. Click the BrowseVault extension icon.

The manifest also registers Ctrl+Shift+Y / Command+Shift+Y as a suggested shortcut for opening BrowseVault.

This extension has no runtime dependencies.

npm run validate
npm run check
npm test
npm run icons
npm run store:media
npm run check:store-media
npm run package
npm run verify:package
npm run release:ready

npm run validate also enforces the current trust baseline: exact manifest permissions, no host permissions, no optional permission prompts, no content scripts, no web-accessible resources, no external extension messaging surface, no remote source URLs, no source-level network APIs or dynamic code loaders, and no default chrome_url_overrides replacement of Chrome history.

npm run check verifies JavaScript syntax, static JavaScript imports, stylesheet imports, extension-page module script and stylesheet paths, manifest and extension UI locale message coverage, manifest file paths, privacy/permission disclosure parity, Chrome QA profile-safety guardrails, shared reference sync, store metadata sync, store media dimensions, direct playbook-compliance evidence, file-size budgets, and folder-density budgets so flat runtime/support and documentation folders stay at 12 files or fewer, while feature folders and feature test folders stay at 15 files or fewer before being split again. File-size debt is reported as warnings, and known hard-limit debt is capped so it cannot grow unnoticed.

npm run store:media regenerates the Chrome Web Store promo PNGs in store-listing/chrome-web-store/media/promo/. It uses Python with Pillow as a development-only asset tool; the generated PNGs are committed and are not part of the extension runtime dependency graph.

npm run check:store-media verifies the committed Chrome Web Store screenshots and promo images have the required dimensions.

npm run package writes dist/browsevault-1.0.0.zip and verifies the final ZIP contents, including packaged manifest paths, locale messages, static import targets, module script targets, and exact freshness against the current source tree.

npm run release:ready is a final release gate for the browser-extension playbook's automated gate, manual target-browser, and screenshot/store-copy review requirements. It intentionally fails until docs/release/manual-browser-qa-checklist.md records passed automated commands, load-unpacked checks, and store-media review evidence for the current Git commit.

Release guardrail tests live in test/scripts/.

Examples:

github site:github.com
host:www.github.com source:chrome-history
title:invoice after:2026年01月01日
category:research
date:2026年06月16日
date:2026年06月16日 hour:14
hour:9-17
visits:>=10 transition:typed
hist* title:resear?
url:docs -youtube
"exact phrase"
regex:github|gitlab

The main vault search is used for archived history management. Quick Open uses the same query text to search current browser sources such as open tabs, bookmarks, downloads, recently closed tabs, and closed windows. Quick Open results are read-only and are not affected by vault delete/export actions. Longer plain keywords use a bounded fuzzy fallback, so common one-character typos such as histroy can still find history; short keywords, phrases, regex, and structured filters stay exact unless they include * or ? wildcards. Wildcards work in plain keywords, exclusions, and title:, url:, source:, or transition: filters.

Date filters use YYYY-MM-DD text fields to avoid browser-specific date input formatting. Use date:, day:, or on: for one local calendar day, and after: / before: for ranges. Bare after:YYYY-MM-DD starts at local midnight, and bare before:YYYY-MM-DD includes the full local day. Use hour:14 for one local hour or hour:9-17 / hour:9..17 for an inclusive local-hour range. Domain filters accept site:, host:, or domain:. Manual category rules are searchable with category: or tag:. Visit count filters accept exact values, comparisons such as visits:>=10, ranges such as count:5..12, and minimum shorthand such as visits:7+. Displayed dates can be switched between system locale, ISO, day/month/year, month/day/year, and year/month/day in Settings.

The Rules tab supports manual domain categories, blacklist rules, and whitelist rules. Category rules label matching domains and subdomains in search results, can be searched with category: or tag:, and are included in BrowseVault JSON archives as rules. CSV and HTML history reports include the derived category label when one applies.

The Rules tab supports manual vault retention cleanup. Enter a number of days, preview how many old BrowseVault records would be cleaned up, then confirm the cleanup. Cleanup uses the same undoable vault deletion path as selected-record deletion and skips whitelisted domains.

The Rules tab also supports duplicate cleanup. It finds active vault records with the same URL and visit time, keeps the richest record, and moves extra duplicates to undoable deletion. It does not delete Chrome history.

The History tab also supports direct current-result deletion for targeted cleanup by query, domain, date, or hour. Enter a query or date filter first, then use Delete Results From Vault to remove matching records only from BrowseVault with undo. Use Delete Results From Chrome only when you also want matching URLs removed from Chrome history; Chrome removes by URL, so that action can affect every native Chrome visit for the matched URLs and cannot be undone by BrowseVault.

Backup and export filenames use the configured prefix and template. The default template is {prefix}-{kind}-{date}. Templates are filename bases without the extension and can use {prefix}, {kind}, {date}, and {time} tokens. The default save mode uses the browser's Downloads behavior; users can switch to an ask-every-export mode that opens Chrome's Save As prompt through the downloads API where the browser supports it.

Backup status is advanced only by the integrity-checked JSON archive export after its self-test verifies checksum, raw row count, and restorable row count. CSV and HTML files are portability/report exports and appear in recent activity without replacing the last restorable backup metadata.

BrowseVault stores its archive locally in the user's browser profile and does not include analytics, ads, tracking scripts, remote code, or automatic page-content capture. Permission details and data-handling behavior are documented in PRIVACY.md.

Open source under the GPL-3.0 license: https://github.com/molodchyk/BrowseVault

The full license text is in LICENSE.

If this extension saves you time and you want to support its development:

Buy Me a Coffee Patreon

BrowseVault and Research Replay should stay separate products at the user-facing level:

• BrowseVault: preserve, search, back up, export, and clean browser history.
• Research Replay: preserve the reasoning trail behind explicit research sessions.

They may later share infrastructure such as URL normalization, local storage, search indexing, backup, and export.