Is Copyfish safe?
Copyfish routes images through a public CORS proxy and sends tab screenshots to a third-party OCR service via a hardcoded shared API key.
When extracting text from images on a page, the extension passes image URLs through cors-anywhere.herokuapp.com, a public proxy not affiliated with the vendor, which receives both the URL and image content. Screenshots of the active tab are then sent as PNG data to ocr.space API servers using a shared API key embedded in the extension's config file. Both transfers occur without user authentication scoped to the individual user.
AI-generated. Findings may contain errors. Those marked Verified have been manually reviewed.
Publishers can request a review.
Findings
Image URLs routed through public CORS proxy for OCR processing
When you ask Copyfish to read text from an image, the extension can route the image URL through the public cors-anywhere.herokuapp.com proxy.
That proxy can receive the image URL and image content before the extension processes it.
The user starts OCR on an image.
Copyfish routes the image through a public third-party CORS proxy before processing it.
| Field | Value | Why it matters | |
|---|---|---|---|
Image URL | https://example.com/images/receipt.png (illustrative) | Reveals the location of the image selected for OCR and may identify the site or page context. | |
Image content | Image bytes for the selected OCR target (illustrative) | Contains the image data being processed, which may include text or visual information from the page. |
Observed during dynamic analysis: the image context menu flow passes the image source URL into the OCR path, and the fetch target is built by prepending the public CORS proxy host to that source URL.