Blackbar

Redact PDFs and screenshots privately, in the browser. Finds SSNs, account/routing/card numbers, birth dates, contact details and custom words, lets you review and adjust, then exports a flattened file where the hidden text is actually gone.

Privacy model: a static site with no backend. Files are processed in the tab (pdf.js for PDFs, tesseract.js for OCR); all assets are self-hosted and a Content-Security-Policy blocks requests to any other origin. Output pages are re-rendered as images with the boxes burned in, so no text layer or metadata survives.

npm install
npm run dev        # local dev server
npm test           # detector unit tests
npm run fixtures   # regenerate fake test documents (needs poppler's pdftoppm)
npm run e2e        # build + browser tests (uses installed Chrome)
npm run build      # static site in dist/

Code: src/detect.ts (detectors, presets), src/extract.ts (PDF text + OCR → character boxes), src/export.ts (flattened output), src/main.ts (UI).