Webhook v3 triggers & skip past state

Hello,
I run into a major issue while using the Webhooks. After triggering the Webhook on condition on column A, I run a n8n workflow that will add a comment on a column B. Somehow the webhook is triggered again. It should not be (as detailed on your documentation “A webhook will only be triggered if the condition transitions from not met to met during a record event.”).

The webhook should be triggered ONLY if the previous condition state of the row was FALSE AND the new state is TRUE (after insertion/update/delete).

The current issue is when we first hit the trigger, the status of the condition was FALSE then it becomes TRUE (correct behavior of the trigger). After inserting something using the automation at the same row but on a different column, the trigger run again, which is not correct. Reminder : To trigger the webhook, the condition should go from false to true.

Here an example with some data on the attached files

Tested in NocoDB SAAS version & 2 Self Host versions (0.264.4)

Can you provide an explanation or a fix please ?

Important note : everything was working fine until few days and was working good on Webhooks v2

Hello @AMH,

Thank you for bringing this issue to our attention. After testing on our end, we were unable to reproduce the reported behavior. To help us investigate further, we would appreciate the following information.

  • can you give a snippet of the update request, especially the Automation value? it’s fine to have the rest of property be redacted
  • is this constantly happening?
  • can you try to make a new table with similar setup and try whether this can be reproducible consistently

If you are able to consistently reproduce the issue with a new table, please provide the step-by-step process so we can replicate it on our end.

We appreciate your cooperation and look forward to resolving this matter promptly.

Hello,

Thanks for your answer.

I created this video so you can see the reproduction and my explanation on what is the issue.

Regards

@AMH Thanks for sharing the video. We were able to reproduce the issue on our side. After investigation, we identified that the bug was caused by an unintended pattern in n8n. We’ll be discussing this with our engineering team, and a fix should be rolled out soon.

We have created a fix here (fix: Apply condition check for prev data in bulk operation hook by pranavxc · Pull Request #12115 · nocodb/nocodb · GitHub) and will soon roll out a bug-fix release. Meanwhile, please verify using the following PR build image:

docker run -d -p 8888:8080 nocodb/nocodb-timely:0.264.4-pr-12115-20250817-0714

Thanks for the team work and the fix !

Hope we get it asap on the SAAS version and self host !

See you soon

1 Like