How are you running AnythingLLM?
Docker (local)
What happened?
The query and the response of the model are not saved in the thread chat when using the thread stream-chat endpoint. Instead, with /v1/workspace/{slug}/stream-chat the messages are correctly stored in the workspace chat even with agent invocation.
Are there known steps to reproduce?
Use /v1/workspace/{slug}/thread/{threadSlug}/stream-chat with an agent invocation.
How are you running AnythingLLM?
Docker (local)
What happened?
The query and the response of the model are not saved in the thread chat when using the thread stream-chat endpoint. Instead, with /v1/workspace/{slug}/stream-chat the messages are correctly stored in the workspace chat even with agent invocation.
Are there known steps to reproduce?
Use /v1/workspace/{slug}/thread/{threadSlug}/stream-chat with an agent invocation.