Notifications
Clear all
How To
5
Posts
4
Users
0
Reactions
6
Views
Topic starter
Describe the problem
When utilizing the Postgres Chat Memory with AI Agent Nodes, I'm unable to incorporate any additional_kwargs, tool_calls, or response_metadata. This occurs even though these fields are added to the database with each record, albeit empty. It would be highly beneficial to have this information, particularly timestamps, for subsequent chat evaluations. My current workaround involves omitting the Chat Memory from the AI Agent Node and manually reading/writing to/from the Chat Memory before and after the call. However, this creates significant overhead to maintain across all sub-workflows that utilize other agents. Simply integrating the Postgres Chat Memory directly into the Agent Node would greatly simplify this process.
Please share your workflow
Information on your n8n setup
- n8n version: 1.62.5
- Database (default: SQLite): PostgreSQL
- n8n EXECUTIONS_PROCESS setting (default: own, main): regular
- Running n8n via (Docker, npm, n8n cloud, desktop app): Docker (GCP)
- Operating system:
Posted : 14/10/2024 10:30 am
Great idea!
That aligns with my requirements as well. I'm looking to incorporate custom data from a tool into a customer profile, and additional_kwargs is a fantastic feature for this purpose.
Posted : 09/12/2024 9:46 pm
Regarding timestamps, a practical solution is to manually set up a chathistory table (the createdat column will be populated automatically):
Posted : 09/12/2024 11:25 pm
Is there a method to incorporate timestamps so that the AI agent can interpret them? My understanding is that the AI agent currently only processes JSON data.
Could you please advise on how to achieve this?
Posted : 23/02/2025 12:24 pm
Are there any new developments regarding this feature?
Posted : 14/03/2025 1:12 pm