Split Pages
🤖 AI Drafted (Not reviewed)
Find the document in a photograph and cut an open book or notebook at its gutter into pages, in reading order; a closed cover or single page is kept whole, an unclear gutter is proposed for review. Sources are never changed.
|
|
| Tool id |
split_pages |
| Category |
transform |
| Uses a language model |
no |
| Needs a generative model |
no |
| Runs over many items |
yes |
| Item handling |
batch |
| Structured output |
no |
| Human-verified |
not yet |
What it reads
| Port |
Type |
Required |
What it is |
Files (files) |
files |
yes |
Photographs of documents. |
What it emits
| Port |
Type |
Required |
What it is |
Pages (output_files) |
files |
— |
The cut pages. |
Parts (parts) |
json |
— |
Each page’s box in its photograph, and the decision for each photograph. |
Options
| Option |
Type |
Default |
What it does |
compression_quality |
integer |
90 |
JPEG/WebP compression quality. |
direction |
string |
ltr |
Reading direction: the order the pages of a spread come out in. One of: ltr, rtl. |
min_confidence |
number |
0.15 |
Below this gutter confidence a spread is proposed for review, not cut. |
output_dir |
string |
— |
Optional output directory. |
output_format |
string |
— |
Format of the cut pages. Default: the source’s own (a JPEG photo stays JPEG). One of: jpg, png, tiff, webp. |
use_apple_vision |
boolean |
yes |
Find the document with Apple Vision’s outline; off, the whole image is the document. |
The prompt it sends
This tool does not send a prompt — it does its work in code.