Privacy policy
Guardian is software you run on your own computer. It has no servers, no accounts and no analytics. The people who make Guardian never receive, see or store your data.
1. Who this applies to
This policy covers the Guardian desktop application, its phone web app (served from your own computer), and this website. "You" means the person running Guardian. "We" means the Guardian project and its contributors.
2. What Guardian stores, and where
Everything Guardian creates is stored on your computer, in a folder named Guardian in your home folder:
- Your goals, settings, notes, journal entries, photos and drawings, workout logs, routine check-offs and library
- Daily briefings (summaries of the sources you choose)
- Sign-in tokens for services you connect, such as Google, in a
secretssubfolder
We have no copy of any of it. Deleting that folder deletes your data.
3. The AI model runs locally
Guardian uses a language model that runs on your computer through Ollama. Your questions, journal, notes, workouts and email summaries are processed on your own hardware and are not sent to any AI provider.
4. Google user data
Connecting a Google account is optional. If you connect one, you use your own Google OAuth client, and Guardian talks to Google directly from your computer.
| Google data | Access | How Guardian uses it |
|---|---|---|
| Calendar events | Read-only | Shows today's and tomorrow's events in your briefing and agenda; answers schedule questions you ask. |
| Tasks | Read-only | Lists your open tasks in your briefing. |
| Gmail | Read-only | Reads the sender, subject and preview of recent unread mail so your local model can summarize it; searches your mail only when you ask it to. |
| Drive file names | Read-only metadata | Lists files you edited recently. |
| One spreadsheet | Create and edit (only files Guardian creates) | Writes your workouts and journal to a "Guardian Log" spreadsheet when you click Export. |
| Email address | Read | Shows which account is connected. |
- Where it goes: Google data is displayed on your computer, and on your phone if you pair one. It is processed by the local model on your computer. It is never sent to us or to any third party.
- No selling, ads or AI training: Google user data is never sold, never used for advertising, and never used to train or improve AI models.
- Storage: briefing summaries that include Google data are stored in your local Guardian folder. Google tokens are stored in
Guardian/secretsand are excluded from Guardian's optional GitHub backup. - Your control: disconnect any time in Guardian → Settings → Google. This revokes the token. You can also remove access at myaccount.google.com/permissions.
Guardian's use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements.
5. Services Guardian contacts for you
To build your briefing and answer questions, Guardian downloads public information from services you configure. These services see an ordinary request from your computer (such as its IP address), the same as if you visited them in a browser:
- News feeds and websites you list, Hacker News, arXiv and Hugging Face
- Market prices (Yahoo Finance), Codeforces and GitHub (public profile data for the usernames you enter)
- Web search (DuckDuckGo, with Bing as a fallback) when the assistant or daily research searches
- Google, if you connect it (see section 4)
Each of these services has its own privacy policy. Guardian does not send them your journal, notes or other personal content.
6. Phone access
If you pair a phone, it connects directly to Guardian on your computer over your local network, or through a private network you set up such as Tailscale. Every request needs a pairing code that only exists on your computer. No relay server of ours is involved.
7. Optional GitHub backup
If you turn on backup, Guardian copies your data folder to a private GitHub repository in your GitHub account. It excludes sign-in tokens, the pairing code, caches and logs. GitHub's privacy policy applies to data stored there.
8. This website
This site is hosted on GitHub Pages and loads fonts from Google Fonts. It uses no cookies or analytics. GitHub and Google may log standard request data under their own policies.
9. Children
Guardian is not directed to children under 13, and we do not knowingly collect any data from anyone. Guardian collects nothing on our side at all.
10. Your choices
- Use Guardian without connecting Google or any other account.
- Disconnect Google, rotate the phone pairing code, or turn off backup at any time.
- Delete all your data by deleting the
Guardianfolder in your home directory.
11. Changes
If this policy changes, the new version will be posted on this page with a new effective date. The history of changes to this page is public at github.com/SoKul-13/guardian-web.
12. Contact
Questions or concerns: open an issue at github.com/SoKul-13/guardian-web/issues.