Inscrutable Red Outline Around Failed Node - flow fails with no error code at all

[Nick1]

No error mssg at all.

No way to examine flow output.

Completely stuck.

This is really not good - no way to debug a flow, lack of inspection at different stages, clunky (super huge) input/output screen.

Also- it’s just burning up my credits - no way to restrict to 1 row of processing on a spreadsheet.

https://www.gumloop.com/pipeline?workbook_id=vXTL3N3KLcSwCN962VS8HR

https://www.gumloop.com/pipeline?workbook_id=vXTL3N3KLcSwCN962VS8HR&run_id=5heiDsazQWeTj38VRRtVcq

PS: I also added a list (for some reason a list is required to process JSON outputs?) and it still fails…

[Gumloop_Bot]

Hey @Nick1! 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.

1. Find your run link on the history page. Format: https://www.gumloop.com/pipeline?run_id={{your_run_id}}&workbook_id={{workbook_id}}

2. Make it shareable by clicking “Share” → ‘Anyone with the link can view’ in the top-left corner of the flow screen.
   

3. Provide details about the issue—more context helps us troubleshoot faster.

You can find your run history here: https://www.gumloop.com/history

[Wasay-Gumloop]

Hey @Nick1 — Appreciate the feedback, it really helps us improve the product.

Regarding your issue:

1. It looks like item #13 failed in the JSON Reader node:
   

   

2. If you inspect item #13 in the run log input/output modal, you’ll notice that the AI response used type instead of the expected company_type key:
   

   

There are a couple of ways to resolve this:

• Bump up the AI model and use explicit prompts to follow the JSON format specified.
• You could wrap the JSON Reader node in an Error Shield, which would catch one-off errors and safely skip to the next input.
• A more robust solution would be to use the Extract Data node. It uses AI to output structured data directly, so instead of generating JSON and parsing it, you can extract just the fields you care about.

You can also limit the number of rows read from the spreadsheet under “Show More Options”:

I’d recommend creating a subflow for everything downstream of the Google Sheet Reader node. That way, it processes one item at a time and can be looped over your spreadsheet. Wrapping the subflow in an error shield will help make the whole flow more resilient. Here’s the tutorial on how to do that.

Let me know if this makes sense and works for you.

I also added some credits to your account so you can continue testing