Scroll · Learn

Install the quiet feed.

ScrollLearn injects spaced-repetition flashcards into Facebook, YouTube, and Instagram feeds. macOS and Windows + Chrome.

↓ Download installer

curl -fsSL https://raw.githubusercontent.com/tasszz2k/scroll-learn/main/install.sh | bash copied

↓ Download installer

powershell -c "iwr -useb https://raw.githubusercontent.com/tasszz2k/scroll-learn/main/install.ps1 | iex" copied

What's inside

More than a flashcard app.

A pocket study system that lives wherever you scroll. Six things ScrollLearn does after you click install.

01 / Learn

Quizzes in your feed

Cards appear between Facebook, YouTube, and Instagram posts every N scrolls. Text, multiple-choice, cloze, and audio formats, scheduled by the SM-2 spaced-repetition algorithm so cards you struggle with come back sooner.

02 / Block

A quieter feed

Hide Reels, Shorts, Sponsored posts, Suggested-for-you sections, and posts from strangers. Add your own keyword filters, organized by topic groups (politics, crypto, drama, whatever is loud this week) and tracked with per-group block tallies.

03 / Shadow

English shadowing

Foundation drills cover all 44 IPA phonemes with curated BBC Learning English clips and a built-in pronunciation grader. Practice mode generates scripts at your level for a four-stage listen / slow / full / blind shadowing loop.

04 / Capture

Capture as you read

Pluck any word or sentence from an allowlisted page. ScrollLearn fetches the translation, dictionary senses grouped by part of speech, and the morphological family. One click promotes a note into a flashcard.

05 / Assist

AI side panel

Ask Gemini to explain whatever you are looking at, summarize a notebook, or generate a quiz from your notes. No API key required for the chat path. Bring your own ElevenLabs or Kokoro key when you want lifelike text-to-speech voices.

06 / Reflect

Stats that respect you

Daily, weekly, and monthly recap. Retention by deck. Annual heatmap. Time on task. Per-topic keyword-block tallies. No leaderboards. No streak shaming.

After clicking download

Open Downloads in Finder and double-click scroll-learn-installer.zip to extract install.command.
Right-click install.command → Open (the first time only; macOS warns about unsigned scripts — click Open).
A Terminal window opens and runs the installer. It downloads the latest release into ~/.scroll-learn/.
The installer opens Chrome at chrome://extensions and a Finder window with the extension folder highlighted. Toggle Developer mode in Chrome, then drag the highlighted extension folder onto the Chrome tab — Chrome loads it.
Prefer Load unpacked? Click it, then press ⌘ Shift . to reveal the hidden ~/.scroll-learn folder (or ⌘ Shift G to type the path).
Copy the extension ID from the ScrollLearn card and paste it back into Terminal.
Reload the extension once. Done. Future updates appear as a one-click banner.

Open Downloads in File Explorer and double-click install.bat.
If Windows SmartScreen warns “Windows protected your PC”, click More info → Run anyway (the first time only; the installer is open source).
A Command Prompt window opens and runs the PowerShell installer. It downloads the latest release into %USERPROFILE%\.scroll-learn\.
If Python is missing, the installer points you at winget install Python.Python.3 — install it, then re-run.
The installer opens Chrome at chrome://extensions and an Explorer window with the extension folder selected. Toggle Developer mode in Chrome, then drag the highlighted extension folder onto the Chrome tab — Chrome loads it.
Prefer Load unpacked? Click it, then paste %USERPROFILE%\.scroll-learn\extension into the dialog.
Copy the extension ID from the ScrollLearn card and paste it back into PowerShell.
Reload the extension once. Done. Future updates appear as a one-click banner.

Updates from now on

The dashboard shows a banner when a new release is out. Click Update now, the new version installs in a few seconds, and the extension reloads. No terminal, no downloads.

Prefer to build from source? Developer install instructions .

Found ScrollLearn useful?

If you find it useful, please star the GitHub repository. It helps other folks discover the project and keeps the work going.