Exporting notes from Skim to Devonthink


Copies notes from PDFs open in SKIM to DEVONthink 2.

For each PDF open in Skim, a correspondingly named folder is created in DEVONthink 2.

  • SKIM notes are exported to the DT folder as RTF records. If the PDF is not in the target folder, it is indexed and linked from DevonThink.
  • Exported note records are placed in the group currently selected in DevonThink, or in a new sub-group if the selected group contains any PDFs other than the one open in Skim.
  • If no active database is selected in DT2, the notes go to the DT2 Global Inbox.


  • Open DEVONThink 2 (and, optionally, select a folder to receive the exports),
  • Open one or more PDFs in Skim,
    Run the script.

DOWNLOAD Ver 0.10 (early draft)

Download page

Hi, thank you very much for this script which I use for neurophysiology literature note taking. However, I would like to turn of the indexing and linking to the pdf which is stored in a different folder. I have looked at your script but cannot make out what line(s) I should mark as comments to stop this function.

The command to look for is indicate.

I am sorry but I have commented out as follows:

set lstDuplic to (children of oLocn where path is strPDFPath)
–if length of lstDuplic < 1 then
–set oPDFRec to indicate strPDFPath to oLocn
–set oPDFRec to first item of lstDuplic
–end if

But, this doesn’t work. Could you please tell me what I should do?

If you download Ver 0.20
You will find that it now has a property near the top:

property pblnIndexPDF : true

if you edit this to:

property pblnIndexPDF : false

  • The PDF will not be indexed,
  • the notes will give the name of the PDF, and any page number, but will not provide a live link.

Thank you for this, but whenever I try to open the scrpt file applescript editor crashes.
Also, if possible I would like the indexed pdf not to be there, but a link to the pdf within the notes is good. The point being that when I do a search on my notes I would like them to be searched but not the “whole” pdf that is also displayed via an indexed link. It somewhat defeats the purpose of extracting highlighted text snippets if the whole paper is also searched because it is indexed completely. Also, I already have indexed all the pdfs in my “papers” folder. But, as I said, I can’t open the script to have a look at it.

The zip for ver 21 contains both text and compiled versions.

The text version should compile under any recent version of OS X.

If the property pblnIndexPDF is set to false, the notes will contain a link to the original position of the PDF file, rather than to an indexing record in DT.

Thank you very much, this works very well.
I am very grateful and do not want to impose, but wonder it there is anyway to turn of the highlighting which is carried over from skim.
Your programme is a major boost to productivity for scientist dealing with a huge literature load.

Formatting details like highlighting can be adjusted by editing the css code in the pstrDefaultCSS property at the top of the text.

You could, for example, edit the line:

highlit {background-color:#FEFC97}

simply changing it to:

highlit {}

Thank you so much for sharing the script!

One thing I am trying to modify the sciprt is expanding the width of the note. Default is about 5in. After I increased the font from 12 to 16, the 5inch is now look quite narrow. I tried to find out where we can change this, but i really don’t know.

Would anyone point out the position where I can change the width to something like ~7in.

Thank you in advance!

You can edit the CSS (at the top of the script, or in an external CSS file if you are using one), to remove the right margin from the quote style.

For example, from:

.quote {color:#333366; margin-left: 30px; margin-right: 30px; line-height: 140%;}


.quote {color:#333366; margin-left: 30px; line-height: 140%;}

Thank you!

Although I would want a defined width, removing the right margin works for me.

Btw, I see the script generates a link (on the URL field) back to the Skim pdf. But the number appears to be lower by one.

For example, when the in-text link page number is 3, the URL link number is 2. But, both links lead to the correct page, which is 3.

Hello houthakker,

  1. In my case, my notes’ font becomes Courier. I would like to change it to something else.
    I searched your script and found only one place mentioning about fonts.

detail {color:#888888; margin-left: 30px; font-family: “lucida grande”, “sans-serif”; font-size:8px;}

I tried differents fonts name instead of “lucida grande” and “sans-serif.” But, it seems that’s not how I should do. Is there an easy way to change the default font?

  1. Personally, when I want to run your script, I find a couple of pdf articles open most of time. In order to import notes from a pdf, I have to close other pdfs. Is there an easy way I can import notes from the front pdf only?

Jy Yun

I’m a newbie when it comes to scripts and DTPO – I’d love to use this one, but when I try to run it as directed via clicking “run” in AppleScript (with DT open and the PDF open in Skim) I get this error message:

“error “Can’t get reference URL of missing value.” number -1728 from «class rURL» of missing value”

What am I doing wrong?

I love this script. I also love this script I found:


I was wondering if there was anyway to put them together so that:

  1. If I highlight text in skim ‘green’ (or any other colour), then they will end up highlighted in that colour in the rtf file (and not all just yellow).

  2. Use a custom url scheme that doesn’t require a pdf alias.

If any applescript experts are out there, it’d be nice to get a “yes, possible” or “not possible” answer. I’ve tried some “tweaking” but I can’t seem to get the script to get the colour of the highlight in skim.

I’d appreciate any help! If not, I think I’ll have to choose between having useful little chunks of information vs keeping certain information (key quotes, summary, methods, references to follow = types of highlights).

Thank you for reading!