Privacy Policy
SavvySearch is a Chrome extension by SavvyWolf AI that turns a webpage you choose into organized, source-attributed research.
Last updated 10/10/2026 · Contact [email protected]
In one paragraph
SavvySearch does nothing until you click Research This Page or Capture As I Scroll on a page. At that moment it reads the text of that one page. With Research This Page, that text and the page's own metadata (URL, title, author and publication date if the page publishes them) are sent right away. With Capture As I Scroll, it also adds text that loads as you scroll, and sends everything once you click Finish Research. If the page links to related pages on the same website (for example a Board or Staff page), SavvySearch first shows them as a checklist; it reads only the pages you leave checked, at most five. Either way the text is sent once to the SavvySearch backend, which forwards them to OpenAI to produce your research summary. The result is returned to your browser. SavvySearch has no accounts, does not track your browsing, and does not collect your name, email or location.
What is collected, and when
Text content of the page you are researching
Including image descriptions (alt text) and the addresses of those images.
When: only when you click Research This Page or Capture As I Scroll on that page.
Where it goes: sent to the SavvySearch backend and to OpenAI immediately (Research This Page) or when you click Finish Research (Capture As I Scroll).
Why: it is the material being researched; image addresses let SavvySearch show a person's photo from the page.
Text of other pages on the same website that you check in the "Also read these pages" list
When: only after you confirm the list.
Where it goes: your browser downloads those pages directly from that website, then their text is sent to the SavvySearch backend and to OpenAI with the page you opened.
Why: you asked for research across those pages.
Your follow-up questions, plus the text of the page(s) you ask about
When: only when you click Ask.
Where it goes: sent to the SavvySearch backend and to OpenAI each time you ask.
Why: to answer the question from those pages.
The research results saved in a project
When: only when you click Combine in a project.
Where it goes: sent to the SavvySearch backend and to OpenAI.
Why: to compare the sources in your project.
Page URL, title, domain, and any author/publication date published in the page's own metadata
When and where it goes: same session and same destination as the page text.
Why: source attribution: showing you where the research came from.
An anonymous random session identifier
When: generated once on your device.
Where it goes: sent with each research request.
Why: rate limiting, so a free service is not abused.
Your IP address
When: with each research request, as with any web request.
Where it goes: backend server logs and rate limiter.
Why: rate limiting and basic abuse prevention.
Pages that build their content with JavaScript
Some websites send an almost empty page and fill it in with JavaScript. To read such a page, SavvySearch asks Chrome for permission to access that one website; Chrome shows the request and you can refuse it. If you allow it, SavvySearch opens the page in a background tab, reads it like the page you opened, and closes the tab. You can remove the permission at any time in chrome://extensions (SavvySearch, Details, Site access).
What is not collected
- No browsing history. SavvySearch cannot see pages you do not research.
- No background or silent collection. Nothing is read when no session is running.
- No account, name, email address, phone number or payment information. V1 has no sign-in.
- No location data, health data, financial data, or authentication information.
- No cookies read or written on the pages you visit.
- No advertising identifiers, and no analytics or tracking SDKs in the extension.
- No selling or renting of data to anyone, and no use of your content for advertising.
Where your data is stored
In the browser. The current session's status, counters, finished result, follow-up questions and the captured page text are kept in chrome.storage.session, which Chrome clears when you close the browser. Pages you add to a project (their research results and captured text) are kept in chrome.storage.local on your device until you remove them, start a new project, or uninstall the extension; they are never uploaded except when you click Combine or ask a question about the project. A single random identifier is also kept in chrome.storage.local.
On the SavvySearch backend. In the default V1 configuration the backend is stateless: it processes a request, returns the result, and keeps no copy. Standard server logs may record request metadata (timestamp, IP address, response status), not page content.
Third parties
- OpenAI processes the captured page text to generate the research output. See OpenAI's privacy policy and their API data-usage terms. Data sent through the OpenAI API is not used to train their models by default.
- Our hosting provider processes requests on our behalf.
No other third party receives your content. SavvySearch contains no analytics, advertising, or tracking services.
Your choices
- SavvySearch only acts when you click Research This Page or Capture As I Scroll. If you click neither, nothing is read.
- In Capture As I Scroll, click Cancel to discard everything captured. Nothing is sent.
- Uninstalling the extension removes its local storage, including the anonymous session identifier.
- To ask about data associated with your anonymous identifier, contact us at [email protected], including the identifier. Note that in the stateless default configuration there is nothing to retrieve or delete beyond transient server logs.
Children
SavvySearch is not directed at children under 13 and does not knowingly collect their data.
Changes
If this policy changes materially, we will update the "Last updated" date and the version published in the Chrome Web Store listing.
Compliance with Chrome Web Store policy
SavvySearch has a single purpose: turning a webpage the user explicitly chooses to research into an organized, source-attributed summary. All data handling described above serves that purpose. We do not sell user data, do not transfer it for purposes unrelated to that single purpose, and do not use it for creditworthiness or lending decisions.
Contact
Email: [email protected]