Recall Search — Privacy Policy

Last updated: July 28, 2026

Recall Search is a Chrome extension that indexes your bookmarks and browsing history locally on your device so you can search them. This policy describes what data the extension handles and how it is used.

Summary

• All search indexing and search queries run on your device.
• We do not operate a backend server and do not collect analytics or telemetry.
• We do not sell your data.
• Optional features may send limited data to third-party services you configure (see below).

Data stored locally on your device

The extension stores the following data in your browser only (IndexedDB and chrome.storage.local):

• Bookmark URLs, titles, and metadata from Chrome bookmarks (if enabled in settings).
• Browsing history URLs, titles, and visit times from Chrome history (if enabled in settings).
• Page content extracted from pages you visit or that are indexed, used to power search.
• Text chunks and on-device embedding vectors generated for semantic search.
• Your search queries, recent searches, and last search results.
• Extension settings (sources, lookback window, filters, theme, etc.).
• Optional LLM API key, if you choose to enable LLM enrichment.
• Optional debug logs, if you enable debug logging in settings.

This data is not synced to Chrome Sync and is not transmitted to the developer.

Chrome permissions

The extension requests permissions solely to provide its features:

• bookmarks — read bookmarks for indexing.
• history — read browsing history for indexing.
• tabs — open or focus tabs when you select a search result.
• storage — save settings and search state locally.
• alarms — schedule resumable background re-indexing.
• offscreen — run the bundled on-device embedding model.
• host permissions (<all_urls>) — run a content script on pages you visit to extract readable text, and optionally fetch pages during background indexing.

Network requests

Core search works offline. The extension may make network requests only in these cases:

• Favicon images — the UI may load favicons from Google’s public favicon service using the site hostname only (no page content is sent).
• Background page fetch (optional) — if you enable “Fetch unvisited pages in background,” the extension fetches URLs from your history/bookmarks to extract content. Requests go directly from your browser to those sites.
• LLM enrichment (optional) — if you enable LLM enrichment and provide an API key, excerpts of indexed page content are sent to the provider you choose (OpenAI or Anthropic) to generate keywords and summaries. This feature is off by default; your key is stored locally and never sent to the developer.

The bundled embedding model runs entirely on your device. The extension does not download models from a CDN at runtime.

What we do not do

• We do not collect, store, or process your data on our servers.
• We do not use analytics, advertising, or tracking SDKs.
• We do not sell or rent user data.

Your choices and controls

You can control what is indexed and stored through the extension settings, including:

• Disabling bookmark or history indexing.
• Turning off page capture on visit or background fetch.
• Disabling semantic search or LLM enrichment.
• Setting URL/title exclusion patterns and purging excluded pages.
• Clearing old history pages or removing all indexed data.
• Uninstalling the extension, which removes locally stored extension data.

Children’s privacy

Recall Search is not directed at children under 13, and we do not knowingly collect personal information from children.

Changes to this policy

We may update this policy from time to time. The “Last updated” date at the top will reflect the latest version. Continued use of the extension after changes constitutes acceptance of the updated policy.

Contact

Questions about this policy: open an issue on GitHub.