lusouldepth-ai/review-notebook — explained in plain English
Analysis updated 2026-05-18
Log a child's schoolwork mistakes by subject, date, and category for later review.
Photograph a mistake and let OCR fill in the record automatically.
Track which mistakes a child has mastered versus which still need review.
Export a printable summary of recent mistakes as TXT, Markdown, or PDF.
| lusouldepth-ai/review-notebook | home-assistant/mobile-apps-fcm-push | ruanyf/nilka | |
|---|---|---|---|
| Stars | 40 | 40 | 40 |
| Language | JavaScript | JavaScript | JavaScript |
| Last pushed | — | 2026-07-05 | 2018-07-26 |
| Maintenance | — | Active | Dormant |
| Setup difficulty | moderate | moderate | easy |
| Complexity | 2/5 | 3/5 | 1/5 |
| Audience | general | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
OCR and speech transcription require a separate Python 3.11 virtual environment and ffmpeg, on top of the main npm app.
This project is a local web app prototype, written in Chinese, that helps a parent keep a record of their child's mistakes from schoolwork, sometimes called a mistake notebook. It runs entirely in the browser and stores everything in the browser's local storage, with no external database or server needed. A parent logs in locally with a phone number or email, and can manage profiles for more than one child. Mistakes are entered manually with a date, subject, category, status, and tags, or they can be filled in automatically by pasting a photo of the mistake for the app's OCR feature to read, using an open source engine called rapidocr. There is also speech to text support through an open source tool called faster-whisper, so a parent could record themselves describing a mistake instead of typing it. Before saving, the app checks whether a similar mistake was already recorded and offers to link the new entry to the old one instead of creating a duplicate. Each mistake can be reviewed, edited, and marked with a status such as not yet reviewed, reviewed, mastered, or needs another look. The app can filter the full list by date range, subject, category keywords, status, or tags, and can summarize weak points such as frequently repeated mistakes, errors that show up across several weeks in a row, or common patterns, though it will say plainly if there is not enough data to draw a conclusion yet. Records can be exported as TXT, Markdown, or PDF files, either as a simple version or a more detailed one, covering a day, week, month, or custom range, and exports can be scheduled to happen automatically on a recurring basis. The app also supports backing up all local data to a JSON file and restoring from it later, asks for confirmation before any risky action like deleting a mistake or restoring a backup, and keeps a log of these sensitive actions. To run it, you start the app with npm start, which serves it at localhost by default, and the OCR and speech features additionally require a Python 3.11 environment with their own dependencies installed.
A local, browser-only web app for parents to log, review, and export a child's schoolwork mistakes, with optional OCR and speech input.
Mainly JavaScript. The stack also includes JavaScript, Python, faster-whisper.
No license information was found in the explanation provided.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
double-check against the repo, no cap.