reword_nerd — Combined Quick Guide
Choose the model profile and rewrite settings that match the chat interface you plan to use.
Add a supported document. Extraction and optional image or OCR processing happen locally in the browser.
Review the extracted text and visual assets, then confirm the document.
Build the package and choose the Runbook, One-shot, or Manual workflow.
Copy prompts into your selected model, or download the ZIP and continue from the standalone workbook.
From TEXT to IMAGE.
Enter the Image portal from the TEXT / IMAGE switch.
Add images or a folder, then click or activate an image card to focus it.
Apply shared settings to selected images using Defaults or Selected.
Review warnings and optional local OCR before accepting any visible text.
Confirm the image set and build the confirmed package in memory.
Review the prompt pairs and download the timestamped ZIP only when ready.
No model runs and nothing uploads; the session stays local in this browser.
Try reword_nerd.