Hey all, I’m having an issue with the generate file node. My flow takes a scanned PDF file of a textbook, goes thru the ask AI node and creates a practice quiz using the textbook info. In my interface I have the user inputting what they want to name the output pdf file. At my Generate File node my flow is only successful IF I have the output link toggled to YES. HOWEVER, it generates a SUUUPER LONG REALLY RANDOM file name. Of course it can be renamed after it is downloaded but I want it to be a better user experience.
When I toggle the output link to NO, and I link the input from the interface text string to the generate node into the file name field, the flow doesn’t pass along any info and my flow fails. I have even tried going into the input settings of the node and toggled the “show as input” toggle for the outlink on, that way the output link can be on YES but I pass it the text string for the user file name input and IT STILL FAILS…this time because of some type of boolean error saying that it was expecting a boolean.
Does anybody have any suggestions OR how do I turn the text string into a boolean output?
Hey @jmigia! If you’re reporting an issue with a flow or an error in a run, please include the run link and make sure it’s shareable so we can take a look.
Find your run link on the history page. Format: https://www.gumloop.com/pipeline?run_id={{your_run_id}}&workbook_id={{workbook_id}}
Make it shareable by clicking “Share” → ‘Anyone with the link can view’ in the top-left corner of the flow screen.
Provide details about the issue—more context helps us troubleshoot faster.
Hey @jmigia – You can use the Split Text node along with a Get List Item node to extract the clean file name from the FileID/File_Name.pdf naming convention.
@Wasay-Gumloop THANK YOU SO MUCH FOR THAT! I really appreciate that you took the time to make that loom video to explain it to me. Thats exactly how I learn rather than reading! Youre AMAZING!!!
@Wasay-Gumloop I adjusted my flow to match your suggestion and I am still getting an error. Just for test purposes I only uploaded one file so it went thru the ‘else’ path. Here is the link
@jmigia – Can you delete the connections to the Join Paths node and then map them again? Its still looking for the File_URL input while that has been switched to the File/File Name. Let me know if that works.
AMAZING THANK YOU AGAIN! It works! It generate the files even with multiple file uploads and the file name is how I wanted it to show. One Question, as you saw I am trying to have it generate a 50-question quiz, but it stopped at 30 and said that it was limited on the output. Is that adjustable?
For the output getting cut-off you can try a different more advanced model like OpenAI o3-mini.
Not related to this but I noticed that the AI output is in markdown (the asterisks and hashtag are formatting syntax), if you enable markdown in the generate file node the output would be better formatted unless you prefer plain text.
I included in the prompt to give it a nice font and bold the question and leave a line space before the first answer choice. Is there a way to clean this up a little bit more?
What’s your current prompt and what model are you using? Can you also share the run link so I can view the inputs/outputs please? You can find the run link on the https://www.gumloop.com/history page or through the Previous Runs tab on the canvas.
Hey @jmigia – I can go ahead and create a ticket to improve the PDF parsing in such cases. In the meantime, would it be possible to use a Google Doc writer node instead? Sorry for the trouble here.
