Skip to content
Listings for the AI era
Sign in

9 October 2026

Your privacy

What we use, what becomes public, and your choices.

Who we are

Lexandry is operated by Intergalactisch, the controller responsible for the personal data described here. Email [email protected] for questions, complaints or privacy requests. You do not need an account. Our business details are on Company & contact.

What we use and why

  • Accounts and listings: your email address, account identifiers, sign-in details, the public name and optional photo of your profile, drafts and submitted content let us provide the account and listing service you request. We use these data to perform our agreement with you. Required sign-in and listing fields are needed to provide those features; browsing does not require an account.
  • Managing a listing: to show that you may manage a listing we use the proof you choose, such as an email address at the listing’s own website, a file on that website or a line in your channel description, and for a review your explanation. To hand a listing over, the current manager gives us the email address of the recipient’s account. We use these data to perform our agreement with you and for our legitimate interest in preventing false claims.
  • Security: session records, sign-in attempts, recovery-code hashes and technical request information help protect accounts and prevent abuse. Our legal basis is our legitimate interest in keeping the service secure. We store recovery-code hashes rather than the recovery code itself. Error logs on our server can contain IP addresses and are used only to find and fix faults.
  • Payments: payment references, subscription status and invoice information let us fulfil your purchase and meet tax and accounting obligations. The legal bases are our agreement with you and applicable legal obligations.
  • Rules and moderation: reports about a listing or profile, our decisions, hidden listings, account bans and requests for review let us apply the Listing rules. We rely on our legitimate interest in a reliable and lawful catalog, or on a legal obligation where a complaint concerns unlawful content.
  • Questions and complaints: your contact details, message, relevant listing and our response help us resolve your request. We rely on our legitimate interest in providing support and resolving disputes, or our legal obligations when handling privacy rights.

Public listings and sources

We receive listing titles, descriptions and source links from contributors. A listing shows the public profile name and photo of whoever added or manages it; that profile also has its own public page listing those listings. Information about people can therefore come from someone other than that person, including material described in linked public sources.

Our purpose is to help people and agents find relevant businesses, professionals and works. For personal information about people other than the contributor, this depends on a legitimate interest in providing that directory and on the person’s rights not outweighing that interest. Public availability alone is not permission for unrestricted reuse. You can object or request a correction using our contact details.

Published listing content and contributor names are visible to anyone, including through our agent interfaces. Drafts, sign-in email addresses and account administration are private. Third parties may copy public listings; removal here does not erase their copies.

Service providers

We use the following services to operate Lexandry. Optional sign-in and payment services receive data when you use those features.

  • Hetzner Cloud hosts our application and database in Nuremberg, Germany, including account data, listings and requests. Data protection.
  • Ploi manages our server, deployments and scheduled database backups, with administrative access to server configuration and operational information. Privacy policy.
  • Cloudflare handles web traffic and security, including IP addresses and request information. We also use its R2 storage for private database backups and a separate deletion record, described below. These stored copies use R2’s European Union jurisdiction. Data processing terms. When many sign-in codes are requested from one network or mail domain, or sign-ins look automated, the sign-in page shows Cloudflare Turnstile before we email a code. Turnstile reads signals from your browser and device to tell people from bots; Cloudflare decides about part of that data itself. Turnstile privacy addendum.
  • Postmark sends our email: sign-in codes, security notices and messages about your listings, payments, management and moderation. It receives the recipient address, message and delivery information. Postmark processes email in the United States. Data protection and retention.
  • Stripe and Link provide checkout and Managed Payments, including payment processing, tax handling, fraud prevention, disputes, receipts and payment-related support. They receive payment and billing details, transaction and subscription information, relevant support communications and references linking a payment to your account and listing. Lexandry stores payment references, subscription status and invoice information; card details are entered with Stripe. Stripe also acts as an independent controller for its own services and purposes such as fraud prevention and legal compliance. Link order management is a Stripe service, separate from your Lexandry account. Privacy policy.
  • Google processes your sign-in when you choose that option, and shares your Google identifier and email address with us. Privacy policy.

Some providers process data outside the European Economic Area, including in the United States. Postmark’s data processing terms, Cloudflare’s data processing terms and Stripe’s data transfer terms describe their transfer safeguards, including European Commission standard contractual clauses where applicable. Contact us for information about the safeguards relevant to your data.

Listing statistics

We record listing views, main-link clicks and supporting-source clicks with their listing and source identifiers, time and a random measurement-session reference. Events are retained for up to 12 months. Searches matching existing catalog tags are counted for seven days; free search text, IP addresses and full user-agent strings are not stored in these measurements. Current authorised makers with an active paid plan can see aggregate Insights for their own management period. Intents requires Extended; previous managers and free contributors cannot access these reports. Their own signed-in visits and recognizable bots are excluded, and assistants that read for someone are reported separately.

Visitor statistics

For every public page that is read, on the website, as Markdown or JSON, or through our MCP tools, we count who read it in one of four groups: people, AI assistants, AI crawlers and other bots. We keep the group, the kind of page and which public listing, category or period it was, the format, and for machines a name from a fixed list (such as GPTBot or Googlebot) based on what the visitor says about itself. For people we also keep where a visit came from as one of four classes (straight here, a search engine, an AI answer, another website), with the name of the search engine or AI service; the address of another website is not kept. We do not store IP addresses, full user-agent strings, free search text, accounts or locations.

Single page reads are kept for 24 hours, linked to the random measurement-session reference, so the operator can see who is reading right now and the pages of one visit. After that only daily totals remain, kept for up to 12 months. Only the Lexandry operator can see these statistics; the operator’s own visits are not counted.

For each public listing we also keep two daily totals for its Insights: reads by AI assistants and AI crawlers under the name from the fixed list, and visits to its page by where they came from, using the same four classes and service names. They hold no person, session or address and are kept for up to 12 months. Only the listing’s current manager on a paid plan sees them, from the start of their own management.

Link checks after publishing

Lexandry opens the links that public listings show again from time to time: every main link, and the sources of paid listings. Each check records the time, whether the page opened, its status code, the address it ended on, and the page’s own title, description and canonical address. We never keep the page itself. These records are kept for up to 12 months, and always the last two checks of a link; a link no listing shows any more is forgotten after 30 days. The check follows a site’s robots.txt and identifies itself as Lexandry-link-check. A listing’s manager sees what the checks found in their dashboard; the public page shows only the date of the last check and a note on links that did not open.

Search context

We store searches and any explicitly provided intent internally to improve search, with no automatic expiry. An intent also orders the results of the search it comes with: among equally matching listings, those holding more of its words come first. This includes search text, applied filters, result identifiers and linked detail retrievals. Before storing intent and search text, we replace recognisable email addresses, clearly marked phone numbers and IBANs that pass our format and checksum checks. Amounts, requirements and ordinary product URLs are retained. The search itself uses the original search text; an intent orders results in its stored, reduced form. The same reduction applies to older history before it is made available by the updated application, including after backup restoration. This is not anonymisation: names, secrets and indirect identifying details may remain. Keyed comparison values support exact continuation and retries for the 24-hour context lifetime; an hourly task clears them after expiry and resumes after downtime. Search history itself has no automatic expiry. Missing intent remains unknown. A random continuation token links a search journey for up to 24 hours; it is not an account or person identifier. We do not provide raw search history to listing managers. For searches by agents, managers with the required active plan can see which search words stood in their listing and which did not, after the same reduction; never the searches of people on lexandry.com one by one. With each search we keep which assistant made it, as a product name from a fixed list taken from the name the client gives or its user agent, never the raw name, and the model an agent states, only when it reads as a model name. Managers see both as stated; we cannot check them. Explicitly provided intent, after we remove recognisable contact details such as email addresses, phone numbers, web addresses and bank account numbers, can be visible to managers of listings the search reached. Only managers with the required active plan see it, limited to their own management period and observed listing interactions. Lexandry does not run an AI to rewrite or certify this text, and does not detect names. Intent provided before 28 September 2026 is never shown. A manager can report a text; we review every search context behind it and can delete it. Leave out personal or confidential information. Contact us using the privacy contact above to request targeted deletion.

Cookies

The application uses a necessary session cookie for sign-in and security, with a seven-day lifetime. There are no advertising trackers. A separate first-party measurement cookie expires after 30 minutes of inactivity. It groups listing views, link clicks and page reads into visits; it does not identify unique people and is separate from your sign-in cookie. Signing out ends the current session. Your browser can block cookies, but account features then may not work.

How long we keep data

Retention depends on what the information is needed for:

  • Account details and private drafts are kept while needed to provide your account and manage your listings. You can delete your account yourself under Security. That signs you out everywhere and removes your email address, public name and photo, ways to sign in, drafts, privately saved content and the text of settled questions. Open questions are kept until handled, then their text is removed, unless still needed for an ongoing investigation or legal obligation. Public listings you added or manage stay in the catalog, credited to “A former member”; payment and invoice records stay as described below. You can also email us to delete data.
  • Published listings stay public while publication is active. When a paid period ends without renewal, its published basics remain public as Free unless hidden or restricted. Paid tags and sources are kept privately for their contributor to read, delete or explicitly restore; private drafts do not become public.
  • For applications to manage a listing, we keep a minimal record of how and when authority was established. Extra evidence is removed 30 days after final handling. Minimal records for rejected or withdrawn applications and ended management are kept for 90 days; the minimal decision record remains while management is active. A genuinely open investigation can require targeted longer retention.
  • Necessary payment and invoice records are retained for accounting and applicable tax obligations: generally seven years under Dutch law, or ten years for records covered by the EU One Stop Shop schemes where applicable. Specific records may also be needed for fraud investigations or disputes, with their continued need reviewed. Account deletion queues removal of the associated Stripe customer profiles and saved cards once payment processing is settled; failures are retried or reviewed. This does not delete financial history that must be retained, Stripe’s records for its own legal obligations, or a separate Link account.
  • Sign-in codes become invalid after ten minutes, and sessions become invalid when they expire. Periodic cleanup removes sign-in attempt records once they have been expired for 24 hours, expired sessions, sign-in attempt counters older than about a day, security event records older than 90 days, and hourly counts of sign-in limits and checks, which hold no address or network, after 90 days. Cleanup runs every minute while the application is running, independently of email delivery, and resumes when it restarts. Records of a hidden listing or an account ban are kept while the measure applies and afterwards where needed to prevent repeated abuse.
  • Correspondence and requests are kept until the matter is resolved, and afterwards only where needed for a related dispute or legal obligation.
  • Our server’s web, application and database error logs are configured for 30 days of retention, with daily cleanup even when no new errors occur. A server or cleanup failure can delay removal until cleanup resumes. Error messages can contain IP addresses or other personal information relevant to a fault.

You can ask us to delete personal data through Company & contact. Some information may need to be retained to meet legal obligations or handle a dispute.

Postmark states that message content and activity are retained for 45 days by default. Delivery suppression records, such as bounced addresses, are retained separately for longer.

Backups and deletion after recovery

Database backups help us recover the service after a failure. They can contain account details, listings, requests and other information held in the database at the time of the backup. Our regular backup schedule runs daily and is configured to keep 30 recovery points. This is not a fixed 30-day deletion period: failed backups can leave older recovery points in place. Separate copies made during migration or recovery, and local copies left after a failed upload, need separate review and removal when no longer needed for that work.

Deleting data from the working database does not immediately remove it from existing backups. To prevent deleted data from returning to use after recovery, we keep a separate record of accepted deletions for accounts, search contexts and correspondence. It contains the type and technical identifier of the affected record and when deletion was accepted, rather than the deleted content. This record currently has no automatic expiry. Before a restored database is reopened, we apply these deletions again. This recovery step does not send another deletion confirmation email. Data retained for a legal obligation or an ongoing investigation remains subject to the criteria above.

Automated processing

Search results are sorted automatically. Lexandry does not use solely automated decisions about you that have legal or similarly significant effects.

Your choices and rights

Under the GDPR, you can ask for access, correction, erasure, restriction and, where applicable, a portable copy of your data. You can object to processing based on legitimate interests. These rights also apply if a listing is about you and you do not have an account. We may need proportionate information to verify a request. If processing is based on consent, you can withdraw that consent without affecting earlier lawful processing.

We must normally respond within one month; any lawful extension must be explained within that month. You can complain to your local data protection authority, including the Dutch Autoriteit Persoonsgegevens. Send your request using the email address on Company & contact; no account is required.

Changes

The date above identifies this notice. We update it when our processing changes and provide additional information where required.