Privacy

Last updated October 11, 2026

Short version: no accounts, no cookies on public pages, no ads, no trackers, no third-party scripts. We don't store your IP address or connect your searches to you.

Searching

• Public pages set no cookies and load nothing from other sites. Thumbnails are fetched by our server, so your browser never contacts the site an image came from.
• Settings such as the adult-content toggle live in the link (&adult=1) or in your browser’s local storage, not in a cookie.
• The small site icons next to results are fetched by our server (it follows each site’s robots.txt) and cached, so your browser never contacts those sites either. You can turn them off on the settings page.
• To stop abuse we count requests per IP address in memory only, for a few minutes to a day; IP addresses are never written to our database or logs by the application.

Settings keys

The settings page keeps your defaults (theme, site icons, start tab, language, place, sort, filters and which instant answers to show) in your browser’s local storage. There is no account and no email. If you choose to save them, your browser makes a random key (like ÅTR-XXXX-XXXX-XXXX-XXXX) and we store:

• a one-way hash of the key (not the key itself),
• the settings, a small piece of text under 1 KB,
• the day it was saved and the day it was last used (no times).

That is all. We never connect a key to your searches, clicks, IP address or bot identity, and keys are never written to logs. Keys travel only in form posts or an HTTP header, never in a web address. A saving browser checks in at most once a day so its key stays alive. Keys unused for 90 days are deleted automatically, and a key saved but never used again is deleted after 7 days. You can delete yours at any time on the settings page (“delete my key”); it is removed from our server immediately. If you lose the key, we cannot recover the settings.

Instant answers you switch off are skipped entirely, including any lookup they would make (for example weather or prices).

Search statistics

To see what people look for and which searches find nothing, we keep aggregate counts of search terms. For each search we store only: the search text (lower-cased; email addresses and long numbers such as phone or ID numbers are masked), the day (no time), the tab (web, images, …), a guess at the language, whether it came from the website or the API, whether it found anything, and a count. No IP address, cookie, browser or device details, bot identity or time of day. These counts are deleted automatically after 30 days and are visible only to the operator.

To learn which results are useful, result links go through a short redirect on Återstod (/r?t=…) before taking you to the site. For each click we store only: the search text (masked as above), which result was clicked and at what position, the day (no time) and whether it came from the website or the API, as a count. We also count how often each result is shown at each position, per day. No IP address, cookie, browser or device details, bot identity or time of day. These counts are deleted after 30 days. They gently move results up or down for everyone; they never personalize results.

For some searches Återstod shows a short AI-written introduction. To make it, the search text and snippets from the top results are sent to our AI provider, and the introduction is cached under that search text (with the date it was made) so it can be reused. Nothing about you is sent or stored with it.

Server logs

Our web server records the time, the requested path without the query string (so not your search), the response status, size and timing, with no IP address. The server's error log may record an IP address when a request fails. Logs are rotated daily and deleted after 14 days.

Reports and removal requests

• Reports (the ⚑ / Report… link): we store the category, the result and any text you add; no IP address or cookie.
• Removal requests: we store what you submit, including any contact details, to handle the request. Contact details are visible only to our moderators, used only for that request, and deleted 365 days after the request is closed; a record of what was removed (and why) is kept so removals stay in place. For copyright notices, if the person who posted the material sends a valid counter-notice, the law may require sharing the notice with them.

The API

API requests are handled the same way. If a program sends an X-Moonsox-Bot-Id header, we look up that public bot identity at Moonsox to decide its rate limit; we don't store it. A program can also send an X-Aterstod-Settings-Key header to use saved settings; that key is handled exactly like the settings keys above and is never linked to the bot identity or the searches.

Our crawler

AterstodBot fetches public pages and respects robots.txt (details). If a page about you appears in results and you want it gone, see removal requests.

Contact

Sox Glade · [contact email coming soon — use /removal]. See also the terms of use.