clipscribe GitHub Details, Stars and Alternatives

REPOSITORY OVERVIEW

Live repository statistics

★ 1Stars

⑂ 0Forks

◯ 0Open issues

◉ 1Watchers

72/100

OPENREPOHUB HEALTH SIGNAL

Healthy signals

A transparent discovery signal based on current public GitHub metadata.

Recent activity35% weight

Community adoption25% weight

Maintenance state20% weight

100

License clarity10% weight

100

Project information10% weight

100

This score does not audit code, security, maintainers, documentation quality, or suitability. Verify the repository and its current documentation before adoption.

README preview

The Problem

Every time you hear something important in a YouTube clip, TikTok short, lecture, or podcast, you have to:

Pause the video
Open a notes app
Manually type it out
Switch back

Then if you want to pass it to an LLM, you type it again.

ClipScribe eliminates all of that.

How It Works

ClipScribe runs a silent 60-second audio loop in the background. When you hear something worth saving, you tap the floating bubble overlay — without leaving your current app. ClipScribe instantly grabs the last 60 seconds of system audio, transcribes it via Whisper AI, and saves it to your history. Ready to copy, share, or paste into ChatGPT.

Open any app  →  Hear something important  →  Tap the bubble  →  Transcript ready

Features

🎙️ Retroactive capture — grabs the last 60 seconds, no manual recording required
🫧 Floating bubble overlay — one tap without leaving your active app
🤖 Whisper AI transcription — powered by @cf/openai/whisper-large-v3-turbo on Cloudflare Workers AI
🔒 Private by design — audio is processed ephemerally and never stored on our servers
📋 Transcript history — saved locally and synced to the cloud when signed in
🌐 Firebase auth — anonymous use or Google sign-in for cross-device history

Installation (Android Beta)

⚠️ This is a pre-Play Store beta. You will need to allow installation from unknown sources.

Option 1 — Direct APK download:

Visit clipscribe.trymyapp.uk on your Android phone
Tap Download Beta APK
Open the downloaded file and tap Install
Grant audio capture permissions when prompted

Option 2 — Build from source:

# Requirements: Android SDK, JDK, adb, USB debugging enabled
git clone https://github.com/lusa8o8/clipscribe.git
cd clipscribe
.\scripts\build-debug.ps1
.\scripts\install-debug.ps1

Architecture

clipscribe/
├── app/
│   └── src/main/java/com/example/
│       ├── capture/          # MediaProjection, rolling audio buffer, freeze flow
│       ├── overlay/          # Floating bubble service and tap handling
│       ├── transcription/    # Audio prep, Cloudflare Workers AI, result holders
│       ├── auth/             # Firebase auth state
│       ├── storage/          # Local + remote transcript store, sync controller
│       └── ui/               # Compose screens (Home, History, Welcome, Permissions)
├── cloudflare-worker/
│   ├── src/index.mjs         # Worker: /transcribe, /transcripts, /waitlist
│   ├── migrations/           # D1 database schema
│   └── wrangler.jsonc        # Cloudflare deployment config
└── website/                  # Vercel landing page
    ├── index.html            # Marketing page with interactive mockup
    ├── privacy.html          # Privacy policy
    └── vercel.json           # Vercel headers + clean URL config

Backend (Cloudflare Worker)

The cloudflare-worker/ directory contains a Cloudflare Worker that handles:

Endpoint	Method	Description
`/transcribe`	`POST`	Transcribes WAV audio via Whisper AI — requires Firebase bearer token
`/transcripts`	`GET` / `POST`	List or create saved transcripts for signed-in users
`/transcripts/:id`	`DELETE`	Delete a specific transcript
`/waitlist`	`POST`	Join the launch waitlist

Deploy the worker:

cd cloudflare-worker
npx wrangler d1 migrations apply clipscribe_transcripts --remote
npx wrangler deploy

Required Cloudflare bindings:

AI — Workers AI binding (Whisper)
TRANSCRIPTS_DB — D1 database
USAGE_KV — KV namespace for per-user daily quota tracking

Required environment variables:

FIREBASE_PROJECT_ID
TRANSCRIPTION_PROVIDER = cloudflare-binding
TRANSCRIPTION_MODEL = @cf/openai/whisper-large-v3-turbo
FREE_TIER_DAILY_TRANSCRIPT_LIMIT = 5

Firebase Setup

Firebase is used for identity (anonymous + Google sign-in).

Create a Firebase project at console.firebase.google.com
Add Android app with package com.aistudio.clipscribe.vfqmza
Place google-services.json at app/google-services.json
Enable Anonymous and Google sign-in methods

Website (Vercel)

The landing page lives in website/ and is deployed to Vercel.

Import the repo at vercel.com
Set Root Directory → website
Deploy — no build step needed (pure HTML/CSS/JS)

Privacy

Audio loop processing is entirely in-memory on your device
The captured audio is sent to Cloudflare Workers AI only for the duration of the transcription request
Raw audio is never stored on our servers
Android's MediaProjection permission triggers the system recording notification — this is expected behaviour, not a bug

Full details: clipscribe.trymyapp.uk/privacy

Pricing

Plan	Price	Transcriptions
Free Beta	$0	5 / day
Monthly Pro	$1.99 / month	Unlimited
Annual Pro	$9.99 / year	Unlimited + best value

Join the waitlist at clipscribe.trymyapp.uk.

Contributing

ClipScribe is early-stage and feedback-driven. If you run into a bug, want a feature, or have ideas, email me at lusa@trymyapp.uk or:

Open an issue
Fork the repo, make your change, and open a PR