# Your data

Updated November 1, 2026.

What you can see, change and delete yourself, how to ask for the rest, and what happens when you connect clickd to an assistant.

## How do I delete my data from clickd?

Sign in at [your account page](https://clickd.com/account), choose Delete everything at its foot, and type your email to confirm. Your account, its key, every agent connected to it, the sites you claimed, your queue and your log of agent calls are deleted at once. Reports on public sites stay, since they describe pages anyone can read; a site’s owner can have those named by a code or removed, as below.

## What you can do yourself

| What | Where | What happens |
| --- | --- | --- |
| See what is yours | Your account page | Your key, the sites you claimed, the agents connected to you and the changes waiting for your answer |
| Disconnect an agent | Disconnect, beside it | Its access ends at once, and it must ask you again to come back |
| Make a new key | New key | The old key stops working at once; the new one is shown once |
| Choose how changes are made | Beside each site | Automatic, or each change approved by you first |
| Sign out | Sign out | This browser is signed out |
| Delete everything | Delete everything | All of the above is deleted at once |

## Connecting clickd to an assistant

When you connect clickd to Gemini, Claude, ChatGPT or another assistant, you sign in with a code sent to your email and approve what it may do: read reports, manage the sites you claim, or ask for sites to be read again. It gets only what you approved. End it from your account page or from the assistant’s own settings, and its access stops at once.

Each call an assistant makes is kept for 90 days: which tool, about which site, how long it took and whether it answered. Never the text of a page or anything you typed. It is shown only to you.

## What you can ask us to do

Anything you cannot do yourself, ask for through [the form on the page for teams](https://clickd.com/enterprise#book), and say which email or site it is about. clickd answers within 30 days.

- See what clickd keeps about you, or have it corrected.
- Remove an early access email, a session request, or notes your agent sent about its work.
- Have a site you own named by an anonymous code on public pages instead of its address.
- Have a site you own and its reads removed from clickd.

## A site you own

To keep clickd from reading it at all, add `User-agent: mttr` and `Disallow: /` to its `robots.txt`; it is obeyed on the next read, as [How clickd reads a site](https://clickd.com/docs/reads#keep-out) says. To have its reads named by a code or removed, ask as above.

## How long things are kept

| What | Kept |
| --- | --- |
| Your account, key and claims | Until you delete them |
| Agent calls | 90 days |
| Sign-in codes | Minutes, until used or expired |
| The log of requests to clickd.com | 14 days, with no address written whole |
| Reads of public sites | While clickd runs, unless the site’s owner asks for them to be removed |

> **The whole of it.** What clickd keeps, why, and who else sees it is in the [privacy policy](https://clickd.com/privacy). Nothing clickd keeps is sold, rented or used for advertising.

## Sources

- [The OAuth 2.1 Authorization Framework](https://datatracker.ietf.org/doc/draft-ietf-oauth-v2-1/), IETF: How an assistant is given access you approve, and how it ends.
- [Connected Apps Platform policy](https://developers.google.com/connected-apps/connected-apps-policy), Google for Developers: Google’s rules for apps in Gemini and Search, which ask for this page.
- [RFC 9309: Robots Exclusion Protocol](https://www.rfc-editor.org/rfc/rfc9309), IETF: How a site’s robots.txt keeps a crawler out.