Privacy

Last updated 29 August 2026

dripdex takes an HTML page and gives back a link with a password on it. This page says what that involves us handling, how long we keep it, and what we will never do with it.

It covers https://dripdex.ai, the pages served at https://dripdex.app, the /v1 API, the MCP connector and the command-line client. It is written to be read, not to be survived.

The short version

No account is needed to publish, and nothing here asks for your name.
An email address is optional, unverified, and used only to reach you.
Your pages are kept until you delete them. Deleting one is immediate.
We keep ordinary request logs, including the calling IP address, for about a week.
Our product counters are aggregate. They carry no page identifier and no viewer address, and that is enforced in the code rather than promised in a document.
There are no ads, no ad networks, no third-party analytics, and no trackers — not on our pages, and not on yours.
We do not sell your data, and we do not use what you publish to train models.

What we handle

What	Why we have it	How long
The page you publish — your HTML, CSS, images and fonts	It is the product. We cannot serve a page we do not store.	Until you delete it. Encrypted at rest, and never reachable without the password.
Earlier versions of a page you have updated	So an update is not a one-way door.	The most recent 3. Superseded bytes leave storage within about 30 days.
Page details — id, slug, title, size, the sanitizer's report, timestamps	To list, serve and meter your pages.	With the page.
A hash of the password (argon2id)	To check a password without being able to read one.	With the page. The password itself is never written down anywhere.
Your account token — an opaque `dd_live_…` string	It is the whole account. It is not attached to a name, and we cannot tell you who you are.	Until the account is deleted.
An email address, only if you give one	To note a claim against an upload and to send a sign-in link. Never required, never verified, never marketed to.	Until you ask us to drop it.
Cookies — three of them, described below	To remember this browser's uploads, to keep you signed in, and to keep a reader from retyping a password on every click.	A year, a session, and about a week respectively.
Request logs — method, path, status, timing, and the calling IP address	To debug faults and investigate abuse. Every service on the internet keeps these.	About a week, then deleted automatically.
Rate-limit counters, keyed on IP address	To stop one caller consuming the service.	They expire on their own inside each limit's window.
Product counters — how many pages published, how many unlocked	To know whether the thing is used.	Up to 15 months. Aggregate only: no slug, no address, nothing that names a page or a person.
Queued email	To hand a sign-in message to the mail provider reliably.	About an hour, then deleted automatically.

The three cookies

dd_drop — set on your first upload, kept a year, HttpOnly. It ties this browser to the pages it has dropped so the box can list them. Clearing it makes the browser forget the account; the pages keep serving.
A signed session cookie, only if you sign in with a magic link.
A page-scoped reader cookie, set after someone types the correct password. It names one page, so it grants nothing anywhere else, and it lapses after about a week.

All three are first-party and strictly functional. There are no advertising cookies and no third-party cookies on this service, so there is no consent banner to click past.

What happens when someone opens your page

They ask us for it, they type the password, and we serve it. That is the whole transaction, and the interesting part is what is not in it.

A published page cannot load anything from another origin. No remote images, no webfonts, no external stylesheets, no scripts of any kind — they are stripped when the page is ingested. So opening a dripdex page does not report the visit to an image host, a font CDN, an analytics vendor or anyone else, because there is nobody else in the request. We see it. Nobody we did not tell you about does.

This is the reason for the no-external-files rule, and it is enforced in the sanitizer rather than requested in the docs.

Publishing through an AI assistant

If you publish through the MCP connector, we receive exactly what the assistant sends us: the page, and optionally a title and password. We do not receive your conversation, your other files, your prompt, or anything about the assistant's user. We do not ask for them and there is no field in which they could arrive.

One thing worth knowing, because it is about your privacy and not ours: the account token comes back in the result of your first publish, which means it lands in that assistant's context and may be stored in its history. That is a credential. Treat it the way you would treat any other one that has passed through a chat window.

What we do with it

We use what is described above to run the service and to keep running it: serving your pages, checking passwords, enforcing limits and quotas, preventing and investigating abuse, diagnosing faults, keeping backups, understanding usage in aggregate, communicating with you about the service, complying with law, and anything else reasonably necessary to operate, secure, support and improve dripdex. We may do any of that without asking first.

We can technically read what is stored, and we may look when there is a reason to: an abuse report, a fault we are chasing, or a lawful demand. We do not browse published pages for entertainment, and there is no internal tool that makes that convenient.

What we do not do

We do not sell, rent or trade your data, and we never will as a business model.
We do not share anything for advertising or marketing purposes, by anyone.
We do not use the pages you publish to train or fine-tune models, ours or anyone else's.
We do not run third-party analytics, ad networks, session recorders, chat widgets or trackers on any page of this service.
We do not build a picture of who read what. The counters that could have supported one deliberately do not carry the dimensions it would need.
We do not put page identifiers or IP addresses into metrics.

Who else touches it

Amazon Web Services, and effectively nobody else. Compute, storage, the database, logs, the CDN and outbound email all run on AWS infrastructure, in the United States, under their terms. Our domain registrar and DNS provider see what any DNS provider sees.

There is no analytics vendor, no advertising partner, no CRM, no support desk and no third-party script anywhere in this service.

Beyond that, we may disclose information when we believe in good faith that it is required by law or legal process, needed to enforce our terms, or needed to protect the service, our users or anyone's safety — and we may transfer it as part of a merger, acquisition or sale of the service, in which case this policy travels with it until it is replaced.

How long we keep it

The table above gives the working answer for each thing. Beyond it: we may retain what we reasonably need for security, operational, accounting, dispute and legal purposes for as long as we consider necessary, and copies can survive in backups and logs for a period after you delete something. Deletion removes a page from service immediately; it is not a guarantee that every byte has already been overwritten everywhere.

Abuse, and pages we take down

We may disable or delete any page, and any account, at any time, with or without notice, if we believe it is being used for phishing, malware, harassment, fraud, or anything unlawful — or if it is simply over quota. We do not owe an explanation before acting, though we will usually give one.

To report a page: abuse@dripdex.ai. Send the URL. You do not need the password, and please do not send it.

Your choices

Delete any page, from the drop page or through the API, the connector or the CLI. It stops serving for everyone immediately.
Rotate a password if you have shared it too widely. The old one stops working and current readers lose access.
Clear your cookies. The browser forgets the account. Your pages keep serving; you keep the token if you saved it.
Give no email. Nothing requires one, and a page published without one is not a lesser page.
Write to us at privacy@dripdex.ai and we will help where we reasonably can, including deleting what we can identify as yours.

Anonymity cuts both ways. Because an account is a token and nothing else, we usually cannot tell whether a request about "my data" comes from the person who published it. If you want us to act on your pages, send the token or something that proves the page is yours. Without one, there is nothing we can safely do — and a service that took your word for it would be a service that took anyone's.

Where it runs, and who it is for

Everything runs in AWS in the United States. If you are somewhere else, using dripdex means your data is processed there.

dripdex is not for children. It is not directed at anyone under 13, we do not knowingly collect their information, and we will delete it if we learn we have.

Changes

We may change this policy at any time. The version on this page is the one that applies, and the date at the top moves when the substance does. Continuing to use dripdex after a change means you accept it. We are a small service and we are not going to email you about a wording fix.

Contact

Privacy
privacy@dripdex.ai

Abuse
abuse@dripdex.ai

What the service actually does, in detail: the manual.