Day 1 Privacy Policy
Last updated: 2 October 2026
Day 1 is made by Kitchen Labs. You paste a public GitHub repository and it makes a short slide deck that explains the code to a new developer. This page says, in plain English, what happens to your information.
The short version
- Day 1 only reads public GitHub repositories. The name you type, the repository's public files and the deck made from them are what it handles, plus a short-lived count per IP address to keep the free service fair.
- A short summary of the repository is sent to an AI service to write the deck. Nothing about you is sent with it.
- Your decks are saved in your own browser, not on our servers. A share link carries the whole deck inside the link.
- There are no accounts, no ads, no cookies and no tracking.
What we handle, and why
The repository you ask for. When you tap Make my deck, your browser sends the repository name or link you typed to Kitchen Labs' server (hosted by Vercel). The server reads that repository from GitHub: its public details (description, stars, language, license, last update), its list of files and folders, the start of its README and a handful of key files, such as package.json or the main entry file. These requests go from our server to GitHub using Kitchen Labs' own GitHub key, so GitHub doesn't see your IP address. Downloads of file contents are made without that key.
Private repositories are refused. If a repository isn't public, Day 1 doesn't read it and answers exactly as if it doesn't exist. Only public repositories can be turned into decks.
What the AI sees. To write the slides, our server sends a third-party AI service a short summary of the repository: its name and branch, description and topics, how many files it has and in which languages, a map of its main folders, the start of the README (with images and links removed), and the first part of up to seven key files. It never sends the whole repository, and it sends nothing about you: no IP address, no browser details, no name. The AI service processes this under its API terms: our AI provider does not use it to train its models, and may keep it for up to 30 days for abuse monitoring. If the AI is busy or slow, Day 1 builds a simpler deck from the files alone, without the AI.
A short-lived cache. So the same repository isn't read and written up twice, our server remembers, in its working memory only:
- what it read from GitHub, for 10 minutes;
- finished decks, for up to 24 hours, keyed by the repository and its current version.
These are public facts about a public repository, not about you. They are never written to a database, and they vanish sooner if the server restarts.
A limit per visitor. To keep the free service fair, the server counts how many decks each IP address has asked for in the last 10 minutes (6 are allowed). That count is kept in the server's working memory for those 10 minutes and then forgotten.
Hosting logs. Like any website, our host Vercel records basic request logs (for example the page or API address requested, the time, and your IP address and browser type), which it keeps for a limited time. Our server also writes short notes to those logs for fixing problems: how much text the AI read and wrote for each deck, and, when a deck comes from the cache or something goes wrong, which repository was asked for and what failed.
What stays in your browser
- Your decks. Every deck you make, and every shared deck you open, is saved in your browser's local storage (up to the 20 most recent), so you can open it again under Your decks. It never leaves your device unless you share it. Delete one with the bin icon next to it, or clear this site's data in your browser settings to delete them all.
- Light or dark mode. Your choice is remembered in your browser.
Share links
Share link puts the whole deck, compressed, into the part of the link after the #. Browsers don't send that part to any server, so opening a share link doesn't send the deck to us. Anyone who has the link can see the deck, so share it only with people you'd show the deck to. Wherever you paste the link (a chat app, an email) keeps its own copy under that service's rules.
What we don't do
No accounts or sign-in, no advertising, no analytics or tracking tools, no cookies, no selling or sharing of data. We don't ask for your name, email or location.
Accuracy
Decks are written by an AI from part of a repository, so they can be incomplete or wrong. Check anything important against the code itself.
Children
Day 1 is not directed at children under 13 and collects no personal information from them.
Changes and contact
If this policy changes, we'll update the date above. Questions? Send them through our contact page.