Privacy.

Who is responsible for this site

Clean Language is published by Jeff Posluns, in Canada, who is responsible for how Clean Language handles personal data. He handles all data protection and privacy enquiries himself, rather than through a separately designated data protection officer, at [email protected]. We do not publish a postal address.

What reaches us when you visit

You fill in nothing and upload nothing by reading the site, because there is no form here. Unless a cache between you and the site serves the page, your browser sends a request to cleanlanguage.ai, and Cloudflare receives it while delivering the site. That request carries your IP address and the address you requested, and it may carry a query, your browser's user-agent and referrer headers, and routing and security signals. Cloudflare records when it arrives. From those, Cloudflare can derive attributes such as country, browser, operating system, and device type. It uses them to deliver the pages, to keep the site secure, and to produce traffic analytics.

Those analytics, together with the sampled security events described below, are the only views we have of requests to this site. The analytics are counts and derived attributes, including a unique-visitor figure that Cloudflare calculates from IP addresses, so they are not purely anonymous. The security events show individual addresses and requests. GitHub separately shows us its own repository traffic figures and a download count for each release asset. We do not use them to identify you, and we do not combine them with anything else. The site shows no consent banner and needs no account to read. Cloudflare still processes each request as described above.

Cookies and storage

The site's own code sets no cookies and includes no client-side analytics. It uses no local storage or session storage in your browser. The site's JavaScript, which handles the copy buttons and the install picker, makes no network calls and writes nothing to cookies, Web Storage, or IndexedDB. The copy button tries to write the text you are looking at to your clipboard when you click it, and selects the text for you to copy by hand if that fails, and the install picker rewrites the address in your browser's history when you choose an AI. You can read both scripts in the published source.

Cloudflare, which serves the site, processes request data such as IP addresses for delivery, security, and aggregate traffic analytics, and may set a security cookie. Cloudflare describes that cookie as one that helps distinguish human visitors from automated traffic, and not as an advertising cookie.

Who else is involved

• Cloudflare hosts the site, serves the Markdown instruction files, and runs the email routing behind the privacy address described below. Its Pages integration also reads the GitHub repository to build the site, so it receives the repository's contents and the deployment metadata attached to each push. All of that falls under its own privacy policy (opens in a new tab).
• GitHub serves the skill archive and the source code, and runs the automated checks on this repository, which process the pushed content along with commit and actor details and keep run logs. Following a GitHub-hosted download or source link makes a request to GitHub, and downloads are redirected to content delivery servers GitHub operates. Both keep their own records under GitHub's privacy statement (opens in a new tab). If you report a problem by opening a GitHub issue, your GitHub account name and everything you write there are normally public, and the maintainer can read it as the repository owner. The same applies to pull requests, comments, reviews, and the authorship details attached to any commit you contribute.
• The sites we link to. The clickable links on these pages go to external sites including GitHub, GitHub Docs, LinkedIn, Creative Commons, Cloudflare, and grclibrary.ai. Following any link tells that site you visited it, under that site's own terms.

We do not sell personal data, disclose it for advertising, or pass it to data brokers. Data does reach the companies above, because delivering a website and an email address requires it, and it reaches any site you choose to click through to. Anything you post publicly on the GitHub repository is public by design. There is little for us to disclose. We do not keep correspondence beyond answering it, and the request logs and analytics sit with Cloudflare and GitHub.

If you write to us

Writing to the privacy address is the only private channel this site offers, and using it means choosing to give us personal data. Your message passes through Cloudflare Email Routing, which records routing events such as sender, recipient, subject, and delivery outcome, and delivers it to a Cloudflare Worker behind the privacy and data protection interface that Jeff Posluns runs across his projects. It does not go out to a separate email provider. We use what you send only to read and answer it. We add you to no list, and we do not keep your message beyond dealing with it.

A message can contain whatever its sender chooses to put in it, including sensitive details we never asked for. Please send only what your question needs.

Where data is processed

Cloudflare and GitHub operate globally, so your request data, and the messages you send to the privacy address, may be processed on servers outside Canada, including in the United States. We hold no copy of the request logs ourselves, and we do not keep your messages once we have answered them. Each provider publishes its own terms and safeguards for the service it runs, linked above, and we have not put separate transfer agreements in place beyond the terms those services are supplied under.

How long data is kept

The website code stores no visitor data. Cloudflare and GitHub retain their own logs and analytics for their own periods, set by them under their own policies. Cloudflare also keeps email routing records, which show who sent what to which address and when, for a period it sets; those outlive the message itself and we can look at them. We do not set, extend, or control those periods. Some records do sit inside our own accounts: the run logs from the automated checks, and the deployment history and build records Cloudflare Pages keeps for the site. We can read those, and delete them where the provider allows it. We do not keep the messages people send to the privacy address beyond reading and answering them. The name and email address attached to a commit are the ones the contributor published through GitHub themselves, and git history is distributed by design, so they travel with every copy anyone makes of the repository, ours included. Public contributions to the GitHub repository, such as issues, comments, and pull requests, are hosted by GitHub under its retention rules; issues and comments can be edited or deleted, while commits already pulled into other people's clones cannot be recalled.

The Clean Language skill itself

The Clean Language skill is a small set of files: Markdown instructions and references, a configuration file, licence and notice files, and the icon in SVG and PNG. It contains no executable code and makes no network calls, and it runs inside your own AI assistant. It carries no telemetry and sends us no prompt and no response, so we never see your writing through it. What the AI provider you installed it in does with your text is governed by that provider's own terms, which we neither control nor describe here. The same holds if you skip installing and paste the instructions or the demo prompt into an AI instead: your writing goes to whichever AI you use, not to us. Who else sees it depends on that assistant, how it is configured, and what it is connected to: a hosted assistant sends it to its operator, and a model running on your own machine keeps it there only if it is genuinely offline and calls no outside service. If a prompt asks the AI to fetch the standard, that request goes wherever the prompt points it: the demo prompt names GitHub, while the Copilot setup names a copy on this site, which reaches Cloudflare.

Your rights

Depending on where you live, data protection law may give you rights over personal data about you, including the rights to see it, to correct it, to have it erased, to restrict or object to how it is used, and to complain to a supervisory authority. Where any processing rests on your consent, you can withdraw it.

Send any request about this site to [email protected]. We keep little to act on, because we do not retain correspondence beyond answering it. We do not control the request logs Cloudflare and GitHub keep and we cannot delete them, though Cloudflare's dashboard does let us look at sampled security events. For those we will tell you what we know and what we cannot do, rather than promise a result you would not get.

Records inside our own accounts, such as check run logs and deployment history, we can reach. Where you have dealt with GitHub directly, by downloading a release or opening an issue, GitHub answers for its own platform records. Issue content also reaches us as the repository owner, and we answer for what we do with it. We will reply as promptly as we can, and within any period the law that applies to you requires. In Canada, the competent authority may be the Office of the Privacy Commissioner of Canada or a provincial or territorial privacy authority, depending on the law that applies and the practice concerned. Elsewhere, where the law that applies to you gives you that right, you can complain to the competent privacy or data protection authority.

Other privacy points

• Data from other sources. We buy no data about you and append nothing from data brokers. Apart from the correspondence and public contributions people choose to send, what we receive from others is the traffic analytics Cloudflare derives from requests, the sampled security events its dashboard exposes, which show individual addresses and requests, and the repository traffic and download figures GitHub reports to us.
• Automated decisions about you. We make no automated decisions about you and we do not profile visitors. Cloudflare's automated security filtering scores traffic to identify what is likely automated, as part of delivering the site, and its outcome is whether a request is served, challenged, or blocked.
• Children's data. The site is not directed at children, requires nobody to submit personal data in order to read or download anything, and operates no age check. Reading and downloading still disclose request data to Cloudflare or GitHub, as described above. A child who visits generates the same request data as any other visitor, and anyone who writes to us sends whatever their message contains.

Changes to this page

We publish any change here and update the date below. There is no mailing list or account through which to tell you, so check this page for the current version.

Last updated: 5 September 2026.

How to reach us

For any question about privacy on this site, write to [email protected].