I have a data table called job_information. I keep all the information about the current job on the table. It seems after the AI agent I can’t seem to access the modified table with all the changes that I have made. When I go look at the actual table there are no rows what so ever. What am I doing wrong?
What is the error message (if any)?
I am getting an error called html_raw not supported so I uploaded my work flow as a file.
You have Delete rows, worked in at wrong time, i.e. being triggered at the wrong time or configured to clear the table too aggressively You have a node named Delete row(s) with a filter set to isNotEmpty. This means every time it runs it will delete any row that has data in it.
AI Agent (yours) uses the resume_type_tool, which is configured for an Update operation.
because your workflow deletes the rows at the start (or if the Insert row node fails), the resume_type_tool will have nothing to update
********
What to do:
Ensure your Delete row(s) node only runs at the very beginning of the workflow to clear old data, and verify that the Insert row node successfully finishes before the AI Agent starts
In resume_type_tool, add a Matching Column (such as: Job_Title or Company_Name). This tells the AI exactly which row in the Application_Data table it needs to update
in the Scrape web page node, try changing the format from raw to text or markdown if you intend to pass the description to the AI
In your Edit Fields node, ensure that the fields like requirements and skills which are currently being set as array, are compatible with the Data Table columnsstrong text
Yeah what Haian said is spot on, that Delete row(s) node with the isNotEmpty filter is basically wiping your table every time it runs. The flow is inserting data and then immediately deleting anything that exists which is why you’re seeing an empty table when you check it.
You probably want to either remove that delete node entirely or change when it triggers, like maybe only delete rows that match specific criteria (old jobs, completed jobs, whatever makes sense for your use case) rather than “anything with data in it”. If you need to clear old data before inserting new stuff, you could add a condition that checks for specific job IDs or timestamps instead of just nuking everything that isNotEmpty.
I am confused. My delete rows is in a n if statement. It should run at the very beginning by deleting all the old data. Then it inserts new data. I should have one row after each run.
Ahh nice, glad you found it . Field mismatches can be sneaky — especially when everything else in the workflow looks correct. My earlier suggestion was based on the common issue where Delete row(s) with isNotEmpty clears everything before the update runs but if one of the fields wasn’t aligned with the Data Table schema that would definitely prevent the row from being saved correctly. Happy you got it working!