How yetone/rewind-for-x works
Rewind for X is a Chrome extension for saving posts a user has actually seen on X’s home timeline and finding them later. The main workflow observes and extracts eligible posts, retains them locally, then presents them in a searchable, filterable drawer. Optional local OCR adds recognized image text to search. The graph also includes the documented toolbar entry point and website hosting. Some content-script internals are unsampled, so those connections rely on README descriptions rather than inferred call chains.
Main entry points: offscreen.js Recognition page · content.js Recall drawer · background.js OCR orchestration · index.js Site request handler
Components 17
Timeline capture
- Post observation
content.js - Post extraction
content.js - Local extension data
- Toolbar shortcut
background.js
Recall interface
- Recall drawer
content.js - Search and filters
content.js - OCR controls
content.js
Image text OCR
- OCR orchestration
background.js - Recognition page
ocr/offscreen.js - Tesseract core
ocr/tesseract-core-simd-lstm.js - Language cache
ocr/offscreen.js
Public website
- Static website
site - Site request handler
site-worker/index.js
Other components
- Extension user
- X home timeline
- Language data CDN
- Static asset service
Connections 19
- Extension user → X home timeline: views timeline
README.md - X home timeline → Post observation: shows posts
README.md - Post observation → Post extraction: extracts eligible posts
README.md - Post extraction → Local extension data: saves posts
README.md - Extension user → Toolbar shortcut: opens drawer
README.md - Toolbar shortcut → Recall drawer: toggles on X tab
background.js - Extension user → Recall drawer: opens in page
README.md - Local extension data → Recall drawer: loads saved posts
content.js - Recall drawer → Search and filters: offers query and filters
README.md - Search and filters → Local extension data: searches saved and OCR text
README.md - Extension user → OCR controls: enables recognition
README.md - OCR controls → OCR orchestration: sets OCR preference
README.md - OCR orchestration → Local extension data: reads posts and stores text
background.js - OCR orchestration → Recognition page: sends image request
background.js - Recognition page → Tesseract core: recognizes image
ocr/offscreen.js - Recognition page → Language data CDN: downloads language data
ocr/offscreen.js - Recognition page → Language cache: caches language data
ocr/offscreen.js - Extension user → Static website: visits product site
README.md - Site request handler → Static asset service: serves static resources
site-worker/index.js