N8N data from main workflow not passing trough in subworkflow

[Scales_Ai]

Hey I am struggling to pass data from my main workflow to my sub workflow. The master ai (router) needs to pass a Telegram ID to a sub workflow. When I look at the input from the master ai to the subworkflow I see the correct data coming trough (a telegram ID) but in my sub workflow it says: query: null. What am I doing wrong?

[tridi]

Hi Scales_Ai welcome to n8n comunity

I saw you just need 1 parameter to pass so why you dont make the sub workflow to accept all data?

But if you still want to define the parameter manually, make sure you click this option to refresh

[Scales_Ai]

[Scales_Ai]

[tridi]

I just tried it on my self-hosted instance, and I’m also not having any issues trying it on n8n cloud.

Have you try to remove that tools and make a new one then use the ID of the sub workflow?

[mredodos]

can you share your workflow?

[Scales_Ai]

how do I do that?

[Scales_Ai]

I tried it yes, doesn’t work…

[tridi]

Block all your related nodes workflow and copy it (ctrl+c)

then when you reply, you can paste after click this button

Result:

[Scales_Ai]

When I accept all data. there is nothing coming trough still…

[King_Samuel_David]

you probably want to define what inputs it needs, in the tools node on main workflow, in description so the ai model know what to send to the tool, this would probably help.

[Scales_Ai]

There is a string available that it send now right? But still there is no data coming in in the subworkflow…

[Scales_Ai]

Eventually, I managed to get it working. I first configured the node to accept all incoming data, then inspected the execution logs to analyze the exact JSON payload the AI was sending. I copied this data and used it as mock data by setting it as a variable. This allowed me to simulate the expected input and verify the workflow behavior.

Interestingly, once I unpinned the mock data, the sub-workflow seemed to have “learned” the expected structure, and from that point on, the correct information was passed through consistently. It appears that initializing the sub-workflow with valid mock input helped it recognize and align with the actual data schema going forward.

[mredodos]

in the description normally you need to set a desciptions of what doing th tool not a variable or something dynamic