Privacy Policy

Data Collection

This plugin does not collect or store any personal data.

File Handling

The plugin receives a file reference (for example, a PDF uploaded to your Dify instance) and downloads that file into memory in order to process it. The download targets the URL provided by the Dify runtime (typically an internal Dify file URL). The file is then parsed locally using the PyMuPDF library.

Local Processing

All PDF parsing — heading/table detection, reading-order reconstruction, header/footer removal, and Markdown/JSON generation — is performed locally within the Dify runtime using PyMuPDF. No OCR model or external service is used, and no file content or extracted data is sent to any third-party service.

Data Retention

The file is processed in memory and is not retained after the request completes.

Contact

For privacy-related questions, please contact: fangyong20062006@gmail.com