Overfront Privacy Policy

Effective date : October 27, 2026

Overfront is a browser extension that shows how visible a website is in AI assistants' answers. It is built to know as little about you as possible. This page describes everything the extension sends, stores, and never touches. Overfront is operated by Estaluna, Inc.

What the extension sends to our servers

• The registrable domain (for example example.com) of a site you view or ask to measure. Never the full URL, never the page path, never the page content.
• A random anonymous identifier, generated once inside your browser, used only to enforce the free measurement quota and prevent abuse. It is not linked to your name, email, Google account, or any other identity.
• A connection key, only if you connect the console. Subscribers can connect the extension to their console account to get more measurements. The console page hands the extension a random key, which the extension sends with measurement requests only. We store a hash of the key, never the key itself; it expires after 30 days without use and is removed when you change your password or delete your account. If you never connect, nothing in this list changes.

That is the complete list. The extension has no access to your passwords, form input, browsing history, cookies, or account data. Besides our API, the only other host your browser contacts on the extension's behalf is the favicon service described under Processors, which receives a domain name and nothing we add. The panel's button opens our console in a new tab; the link carries the domain you were viewing so the page can offer to track it, and nothing else. When you reach a free limit (the weekly measurements, or the one free Refresh of a site), the extension opens that same page once, right after your Measure or Refresh press.

When you remove the extension, your browser opens a page on our website that asks why. Its address carries your anonymous identifier and the extension's version number, and nothing else. Answering is optional : if you answer, we store the reason you pick and any text you write, with that identifier and version, so we can read the answer alongside the measurements made from your browser. To stop abuse, our server limits how many answers one IP address can send per minute; the address is not stored.

What the extension never collects

• Full URLs, page contents, or search queries
• Names, email addresses, or any personally identifiable information
• Browsing history — we receive a domain only at the moment you view or measure it, and our API requests carry no cookies or referrer

What stays in your browser

Your preferences (panel position, on-page badge on/off, muted sites), recently viewed scores, and the anonymous identifier are stored locally via browser extension storage. They leave your machine only as described above and are removed when you uninstall the extension.

How we use the data

• To compute and return AI visibility scores for the requested domain
• To cache scores so all users share results without repeated measurement
• To enforce the free quota and protect the service from abuse
• To build aggregate, domain-level analytics about visibility in AI answers. These aggregates never contain personal data — there is none to include

We do not sell data. We do not use it for advertising. We do not share it with third parties except the processors below.

Public pages our server reads

To prepare the Actions tab's recommendations, our own server — not your browser — fetches the measured domain's public robots.txt file, home page, and sitemap, along with public pages that AI answers cited during the measurement. When AI answers in the domain's category cite Wikipedia, it also reads the one Wikipedia page titled with the brand's name, to see whether an article exists. It reads these public pages even when a site's robots.txt asks crawlers to stay away. When you mark an action as done, the extension sends the domain, your anonymous identifier, and the action's identifier, and nothing else.

Console accounts

The console at app.theoverfront.com is used only by people who create an account; the extension never needs one, and connecting it to an account is optional. For an account we store your email address, the name you give or your Google profile name, a hash of your password (never the password itself), when you signed in, and a cookie that keeps you signed in. We use these only to sign you in, run the domains you track, and bill your plan.

If you sign in with Google, we ask Google only for the openid, email, and profile scopes and receive your email address, whether Google has verified it, your name, and your Google account identifier. We use them only to sign you in and to link the sign-in to your account. We do not store Google access or refresh tokens, do not call other Google services on your behalf, and do not sell this data, use it for advertising, or use it to train AI models. Our use of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements.

You can delete your account yourself under Settings, Delete account. Any subscription ends at once, and your account, tracked domains, settings and Cloudflare connection are deleted. We keep a one-way hash of your email so that a new account with the same email does not get another free trial, and the reason you give for leaving, if any, with no link to you. Stripe keeps billing records as the law requires. Measurements of public domains are shared by everyone and stay. For help, write to [email protected].

Cloudflare connection (console)

Console customers can connect their own Cloudflare account, read-only, through Cloudflare's sign-in or a read-only API token. Once a day our server reads, from the zones that match your own tracked domain, the requests made by known AI crawlers (for example OAI-SearchBot, ChatGPT-User, Googlebot, ClaudeBot), identified by their User-Agent. We store only the domain, the hour, the page path, the crawler's name, and a request count. We never ask for or store human traffic, IP addresses, query strings, or the User-Agent string itself. The token is stored encrypted and is never shown back. These crawler counts are kept with no fixed expiry date, so visits can be compared over any period. Disconnecting revokes our access where Cloudflare supports it and deletes the connection and its zone list; the crawler counts already collected stay, aggregated per domain and not tied to your account.

Processors

• Cloudflare hosts our API, cache, and database. Standard server logs (including IP addresses) are processed transiently for security and are not used to identify users.
• Stripe processes payments for paid plans. We send it your account's email address; card details go to Stripe directly and are never stored by us.
• AI engine providers (currently OpenAI) receive only the measurement questions we generate about a domain. They never receive your anonymous identifier or anything else about you.
• DataForSEO fetches Google's AI Overviews and AI Mode answers for the daily tracking in the console. It receives only the measurement questions we generate, with the language and country to search in. It never receives your anonymous identifier or anything else about you.
• Anthropic writes the measurement questions for a domain. From our servers it receives the domain name and the public title, description, and headings of the site's home page; when Anthropic is unavailable, OpenAI receives the same to write them. Neither receives your anonymous identifier or anything else about you, and nothing is sent to them from your browser.
• TypeSafe (Jev) receives, from our servers, the text of public web pages together with a measurement question, to judge how well a page answers it. That covers your site's public pages and the pages AI answers cited. It also receives the questions written for a domain, with the site's name, domain, and a one-line description of what it offers, to check each question before it is used. It never receives your anonymous identifier or anything else about you, and nothing is sent to it from your browser.
• Google's favicon service (www.google.com/s2/favicons) serves the small site icons shown beside each domain in the panel. Your browser loads them as ordinary images, one per domain : the competitors and cited sources our API returned, and the site you are viewing when your browser has no icon of its own for it. Each request carries that domain and nothing we add : no anonymous identifier, no page address, no score, and no referrer. We receive nothing back from it. Google sees these requests the way it sees any image a web page loads from it, including your IP address and whatever cookies your browser already sends to Google. A site icon is a public asset of the site it belongs to. To know which domains that service has no icon for, our server asks it once per domain and remembers the answer; only the domain name is sent, never anything about you. For those domains the panel shows the icon your browser already saved when you visited the site, read on your device with nothing sent anywhere. When there is none, the panel asks our own API, by domain name only, and our server reads the icon the site itself publishes.

Optional permission : on-page badge

The extension works from its side panel with access to the active tab only. If you enable the automatic on-page badge, the browser asks for optional access to sites you visit; this is used solely to display the badge, the domain is processed locally, and you can revoke the permission at any time in your browser's extension settings.

Website analytics

This website, theoverfront.com, uses Cloudflare Web Analytics to count page views. The extension does not: this section is about the site only. Cloudflare states that Web Analytics "does not use any client-side state, such as cookies or localStorage, to collect usage metrics" and that it does not "fingerprint" individuals via their IP address, User Agent string, or any other data for the purpose of displaying analytics; query strings are not logged. What we see is aggregate : how many times a page was viewed, from which referrers, countries, and browser types. We cannot identify or follow an individual visitor, and nothing from the site's analytics is joined to the extension's anonymous identifier.

Retention

Stored scores are shared across users and refreshed as new measurements become available. Measurement history is kept with no fixed expiry date, so trends can be shown over any period, and is keyed by domain, not by user.

Changes

If this policy changes, the new version is published at this address with an updated effective date.

Contact

Questions about this policy : [email protected]