Privacy policy

Effective

Index is a personal library: you save links, notes and files, and it reads and organises them so you can find them later. This policy says what data that involves, who handles it, and how you can see or remove it.

Who runs this service

Index at https://index.intrface.eu is run by INTRFACE j.d.o.o., based in Croatia. INTRFACE j.d.o.o. decides how your data is used and is the data controller under the EU General Data Protection Regulation (GDPR).

Write to basic@intrface.eu with any question about this policy or your data.

What we collect

Your account

You sign in with your Google account. We keep your email address, your name, a link to your profile picture, and the user id our sign-in provider gives you. Only invited email addresses can create an account.

What you save

Everything you send to Index: links, text, notes, and files such as images, audio, video and PDFs. We record when you saved each one and how it arrived (the web app, your phone's share sheet, the Telegram bot, the API, the Obsidian plugin, your forwarding address, or a feed you follow).

What Index works out from it

To make your items readable and searchable, Index fetches the page behind each link and stores its text, title, description, author and images. It also creates transcripts of audio and video, text read from images, labels such as topic, format and language, names of people and things mentioned, links between items, and search data (numeric vectors of your text). YouTube videos are handled differently: see Google and YouTube data.

How you organise it

Collections, worlds, saved views, annotations and research notes you create.

What you do in Index

Which items you open, how long you spend on them, and what you review or set aside. We use this only to order and suggest items in your own Index.

Spaces and shared investigations

Index uses passages from your selected World or library to explore a question. It stores your question, review history, requested reminders and kept findings. Each kept finding records its source, cited passage, interpretation, your name and any thought you add. Your unkept suggestions and library scope stay private.

An investigation owner can grant another registered account access to the question, participants and kept findings, including original links and timestamps. Participants can contribute their own kept findings. Removing a participant also removes their contributions from the investigation. The original items remain in their library. Deleting an original item removes its kept passages from investigations.

Accounts you connect

  • GitHub: your GitHub account id and username, an access token, and the repositories you have starred.
  • YouTube: your channel id and name, access and refresh tokens, and your playlists and liked videos. See Google and YouTube data.
  • Telegram: the chat id and user id of the Telegram account you link to the capture bot.
  • Obsidian: the name of each vault the plugin syncs, the folders you chose, and the plugin's key, stored like the API keys below. See Notes from Obsidian.

Notes from Obsidian

If you install the Index plugin for Obsidian, it sends Index the Markdown notes in the folders you choose: each note's path, text, size and the time it last changed. It sends nothing from outside those folders, and no images, attachments or other files. It skips notes of 512 KB or more, Obsidian's settings folder and the Index folder.

Index stores the text of each note and the tags, aliases, links and web addresses in it. It sorts each note into your worlds and makes search data from it. It does not automatically fetch linked pages or read note images. You can choose to investigate the text in a Space.

Index writes Markdown files into one folder of your vault, called Index unless you rename it: your worlds, the saved items your notes connect to, Spectre digests and knowledge bundles. It keeps a copy of these files on our server so the plugin can fetch them. It changes nothing in your vault outside that folder. If you sync your vault with another service, such as Obsidian Sync or iCloud, these files travel with it under that service's terms.

Forwarded newsletters

If you use the forwarding address Index shows in Settings, mail sent to it reaches Index through Cloudflare (see Who else handles your data). The address is private to your account and starts with a short form of your name.

For every mail it receives, Index keeps the sender's address and name, the subject, when it arrived, and its message id. For a newsletter issue it also keeps the text of the issue and its web address if it has one, and adds the issue to your library as an item. For a mail that asks you to confirm a subscription, it keeps the confirmation link or code so you can open it from Settings. Index does not receive attachments.

Anyone who learns your forwarding address can send mail to it, and that mail lands in your library like a newsletter. You can replace the address in Settings; the old one then stops accepting mail.

Feeds you follow

If you add a site or feed address to a Spectre, our server checks that public feed every two hours. Each time it looks at up to 20 of the newest posts; the first time, by default, only at posts from the last three days. It saves each new post as an item, the same way it saves a link you send, and skips posts that point to video or audio.

API keys

If you create a personal API key, we store a one-way hash of it and its first few characters, never the key itself.

Technical data

Our hosting providers record basic request data, such as IP address, browser and time, to run and protect the service.

How we use it

We use your data only to run Index for you:

  • to store what you save and show it back to you;
  • to read, transcribe, label and link your items so you can find them;
  • to search your library;
  • to keep connected accounts and your Obsidian vault in sync, and to check the feeds you follow;
  • to keep the service secure and fix problems.

We do not sell your data. We do not show ads or use your data for advertising. We do not build profiles of you for anyone else. We do not train AI models on your data.

Our staff read your data only to fix a problem you report to us, for security, or where the law requires it.

Who else handles your data

Index uses the services below to do its work. Each receives only what it needs for its task and handles it on our behalf.

  • WorkOS (sign-in): your name, email address and profile picture from Google, and your session.
  • Vercel (hosts the web app) and Railway (hosts the server, background jobs and database): all data passes through or is stored on their systems.
  • Cloudflare: R2 file storage holds the files you save and images Index fetches for your items, and no YouTube videos, audio or thumbnails. If you use a forwarding address, Cloudflare Email Routing and a Cloudflare Worker receive each mail sent to it, read it and pass it to Index; Cloudflare handles the whole message, attachments included, on the way.
  • OpenAI: the text of your items, including notes from Obsidian, newsletter issues, and YouTube video titles, descriptions and transcripts, and your search queries, to create search data; audio and video files you upload yourself, to transcribe them.
  • Google Gemini API: images you save, to describe them and read their text; the public web address of each YouTube video you save or sync, so Gemini can read the video and return its transcript.
  • TypeSafe AI: the title, description and up to about 6,000 words of each item's text, to label it and sort it into your worlds; for a note from Obsidian, its title, tags and up to about 6,000 words; when you turn on “Best match first” in search, your query with the titles and short descriptions of the top results.
  • When you request assisted research in a Space, the configured classifier and Google Gemini can receive selected passages and the investigation question. Cloudflare handles classification when CLEF is enabled.
  • Apify: links to Instagram posts you save, to read the post.
  • FxTwitter: links to X (Twitter) posts you save, to read the post.
  • GitHub: links to GitHub repositories you save, to read the repository and its README; with your token, to read your starred repositories.
  • YouTube and Google: YouTube video ids and addresses, to read public video details through the YouTube Data API or YouTube's oEmbed service; with your permission, your YouTube account, as described below. When you look at a video in Index, your browser loads its thumbnail and player straight from YouTube.
  • Telegram: if you use the capture bot, your messages and files reach Index through Telegram.

When you save a link, or a feed you follow has a new post, our server fetches that page. Our server also fetches each feed you follow. The site sees a request from our server, not from you.

You can share selected findings with the people in an investigation. Other disclosures are limited to the service providers listed here and disclosures required by law.

Google and YouTube data

Index uses YouTube API Services. By connecting YouTube you agree to the YouTube Terms of Service. Google's handling of your data is covered by the Google Privacy Policy. Index is not made or endorsed by YouTube or Google.

What we access

When you connect YouTube in Settings, Google asks you to give Index the scope https://www.googleapis.com/auth/youtube. With it, Index reads:

  • your channel id and channel name;
  • your liked videos;
  • the playlists you own and the videos in them;
  • for each video: its id, title, description, channel, when it was published and when you added it.

What we do with it

  • Every hour, Index reads your playlists and liked videos and mirrors them into your library. Your liked videos, and each playlist, appear as a collection.
  • When you unlike a video, remove it from a playlist, or delete a playlist on YouTube, Index removes it too, and deletes any saved item that was there only because of the sync.
  • When you ask, Index adds a video to one of your playlists or creates a new playlist. The change waits in a queue and can take a minute to appear on YouTube.
  • Index never posts, comments, rates or deletes anything on YouTube, and never acts on your account unless you ask.

What we store

  • Your access and refresh tokens, encrypted with AES-256-GCM.
  • For each video: its title, description, channel and dates, from the YouTube Data API or YouTube's oEmbed service.
  • Its transcript, which the Google Gemini API writes by reading the public video, and the search data Index makes from that transcript (short passages and their numeric vectors).

Index never downloads YouTube video or audio. It keeps no copies of YouTube thumbnails and no raw YouTube API responses. Thumbnails load from YouTube's servers, and videos play in YouTube's embedded player.

How we keep it current

Index refreshes the details of every YouTube video at least once every 30 days. If a video has been deleted or made private, Index removes its YouTube data, including the title, description and transcript, and keeps only your own record of saving it and your notes.

How we share it

We never sell YouTube data or use it for advertising. Vercel and Railway provide hosting and storage. Google Gemini receives video addresses to write transcripts. OpenAI and the configured classifier use titles, descriptions and transcripts to make videos searchable and label them.

If you request assisted research, Cloudflare and Google Gemini can receive selected excerpts and your question. If you share an investigation, its members can read the excerpts and thoughts you keep there. An agent you connect with library access can read evidence under that account’s permissions.

Index's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

How to remove access and data

  • In Index: Settings, YouTube, Disconnect. Index revokes its access with Google, deletes the stored tokens, stops syncing, and deletes your synced playlists, their collections, and every saved item that exists only because of the sync. Videos you saved yourself stay in your library.
  • In your Google account, at myaccount.google.com/connections. On its next sync Index finds that access is gone and deletes the same data as a disconnect.
  • Deleting your Index account in Settings also revokes the YouTube connection and deletes all of your data.

Our staff read your YouTube data only to fix a problem you report to us, for security, or where the law requires it.

Cookies and local storage

Index sets only what it needs to work:

  • a session cookie from our sign-in provider, WorkOS, that keeps you signed in;
  • short-lived cookies (gh_oauth_state, yt_oauth_state) that protect the GitHub and YouTube connection steps; they last at most 10 minutes and are removed when the step ends;
  • in your browser's local storage, your light or dark theme choice and, if you enter one, the GitHub username you want to browse;
  • a service worker that keeps an offline page and a device outbox. After you sign in, an installed share can hold text, links and files in this browser until delivery is confirmed. Only your signed-in account can send its drafts. You can discard them from Save or clear this browser’s site data.

Index has no analytics, no advertising and no third-party trackers of its own.

When you play a YouTube video in Index, the player comes from YouTube, and YouTube may set its own cookies under the Google Privacy Policy.

How long we keep it

  • Your account and library: as long as your account exists.
  • A saved item: until you delete it. Deleting an item removes it, its files and everything Index worked out from it. The record of the original capture (the link or text you sent and when) stays in your account history until your account is deleted, or until you ask us to remove it.
  • Tokens for connected accounts: until you disconnect the account.
  • Notes from Obsidian: until you delete a note, move it out of the folders you sync, or disconnect the vault in Settings. If the plugin misses a deletion, for example because it was not running, the note leaves Index once it has been missing from the vault for 24 hours. A plugin that still has a valid key sends a disconnected vault again; revoke the key in Settings to stop it. Disconnecting does not delete the files Index already wrote into your vault. Our copy of those files stays until your account is deleted.
  • Forwarded mail: a newsletter issue stays in your library until you delete it. The record of each mail (sender, subject, time, message id, and any confirmation link or code) stays until your account is deleted, even after you delete the issue or dismiss the confirmation.
  • Posts from feeds you follow: saved items like any other, kept until you delete them.
  • YouTube data: refreshed at least every 30 days, removed when a video is deleted or made private, and deleted when you disconnect YouTube or revoke access. See Google and YouTube data.
  • Your whole account: delete it in Settings. Index revokes the access you gave to connected accounts and deletes your account and all your data, including stored files. You can also write to basic@intrface.eu from your account's email address and we will do it for you within 30 days.

Security

  • All traffic uses HTTPS.
  • Tokens for connected accounts are encrypted with AES-256-GCM.
  • API keys are stored only as one-way hashes.
  • Your files sit in private storage and reach your browser only through Index after sign-in.
  • Only invited accounts can sign in.

Your rights

Under the GDPR you can ask us to:

  • give you a copy of your data (access);
  • correct data that is wrong;
  • delete your data;
  • send you your data in a machine-readable format (export);
  • stop or limit a use of your data (objection and restriction);
  • withdraw a consent you gave, at any time.

Write to basic@intrface.eu. We answer within one month. You can also complain to a data protection authority, in Croatia or in the EU country where you live; the European Data Protection Board lists them.

Transfers outside the EU

Several of the services above are based in the United States, so your data may be processed there. These transfers rely on the EU–US Data Privacy Framework where the provider is certified, or on the European Commission's standard contractual clauses.

Children

Index is not meant for anyone under 16, and we do not knowingly hold their data.

Changes to this policy

The date at the top shows when this policy last changed. If we change how we use your data in a way that matters, we will tell you by email or in Index before the change takes effect.

Contact

INTRFACE j.d.o.o., Croatia. Email basic@intrface.eu.