See the pages of a scan

documents_page_images · Reads only · Markdown

The pages of a scanned PDF or a photo of a document as images to look at, up to 10 pages a call.

$0.005 a call

From your agent’s balance. A call that fails costs nothing.

Look at the photo of the signed delivery receipt and tell me who signed for it.

Your agent callsdocuments_page_images$0.005 a call

The example business is Hansen & Voss, a restaurant-equipment wholesaler in Hamburg with 60 staff. The reply is the one the tool really gives.

With these arguments

{
  "file_id": "8f0b2d4f-6c8e-4a0c-9e2b-5b7d9f1b3c6a"
}

What comes back

{
  "pages": [
    {
      "page": 1,
      "image": "data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAAAQABAAD/2wBDAAUDBAQEAwUEBAQFBQUGBwwIBwcHBw8LCwkMEQ8SEhEPERETFhwXExQaFRERGCEYGh0dHx8fExciJCIeJBweHx7/2wBDAQUFBQcGBw4ICA4eFBEUHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh4eHh7/wAARCARMAyADASIAAhEBAxEB/8QAHwAAAQUBAQEBAQEAAAAAAAAAAAECAwQFBgcICQoL/8QAtRAAAgEDAwIEAwUFBAQAAAF9AQIDAAQRBRIhMUEGE1FhByJxFDKBkaEII0KxwRVS0fAkM2JyggkKFhcYGRolJicoKSo0NTY3ODk6Q0RFRkdISUpTVFVWV1hZWmNkZWZnaGlqc3R1dnd4eXqDhIWGh4iJipKTlJWWl5iZmqKjpKWmp…",
      "text": null
    }
  ],
  "page_count": 1
}

Use it when

The agent has to see a document: a handwritten note, a stamp, a signature.

For a PDF that has a text layer, use documents_read_pdf.
For the printed words of a scan as text, use files_ocr.

What it does

The pages of a scanned PDF or a photo of a document (a receipt, a letter, a signed form) as images for you to read yourself, with any text the PDF already has. Up to 10 pages a call; ask for later pages with first_page. files_ocr turns a scan's printed words into text instead ($0.005 a page); for PDFs with text, documents_read_pdf is free.

What comes back

pages: each with its number, an image you can see (a data URL, or an image block over MCP) and any text it already has; and page_count. Every reply also carries cost_usd and balance_usd.

Inputs

Name	Type	What it is
`file_id`Required	string	A file in the store (files_upload or files_list give its file_id). Exactly 36 characters.
`first_page`	integer	At least 1. Default: 1.
`last_page`	integer	At least 1.

The call

Over MCP, call documents_page_images on https://api.conecto.chat/mcp/documents. Over REST:

Terminal

curl -X POST https://api.conecto.chat/v1/tools/documents_page_images \
  -H "Authorization: Bearer $CONECTO_KEY" \
  -H "Content-Type: application/json" \
  -d '{"file_id":"8f0b2d4f-6c8e-4a0c-9e2b-5b7d9f1b3c6a"}'

When it refuses

A refused call costs nothing and says what to do next.

Code	When	What to do
`not_found`404	No file in the store has that file_id.	Find it with files_list, or upload it first.
`unsupported_type`400	The file is not a PDF or an image.	Read text files with files_read_text.
`not_image`400	The image cannot be read.	Upload it again as PNG or JPEG.
`too_large`413	The image is over 60 megapixels.	Shrink it with image_transform first.
`invalid_pdf`422	The PDF cannot be opened (it may be password-protected).	Ask for another copy.
`invalid_input`400	first_page is past the end of the PDF.	Ask for pages it has.
`pdf_locked`422	The PDF is password-protected.	Ask for a copy without a password.
`busy`503	PDFs are being read for other calls right now.	Try again in a minute; nothing was charged.

Give this to your agent.

Set up Conecto from https://conecto.chat/SKILL.md

See the pages of a scan

An example

Use it when

What it does

What comes back

Inputs

The call

When it refuses

More Documents tools

Give this to your agent.