Description
Describe the solution you'd like
Currently, we can only send "text" using either the SDK like so
for event in app.stream_query(
user_id="u_123",
session_id=session.id,
message="whats the weather in new york",
):or using REST
curl \
-H "Authorization: Bearer $(gcloud auth print-access-token)" \
-H "Content-Type: application/json" \
https://LOCATION-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/LOCATION/reasoningEngines/RESOURCE_ID:streamQuery?alt=sse -d '{
"class_method": "stream_query",
"input": {
"user_id": "USER_ID",
"session_id": "SESSION_ID",
"message": "What is the exchange rate from US dollars to SEK today?",
}
}'But the current adk web supports file upload, so I guess the feature is already implemented, right? If the answer is yes, it should be pretty easy to make it available in the SDK and REST. If it's the case, I would like some guidance to implement this, I'll do it asap.
The goal is to use the same app.stream_query(user_id=......) sdk or REST to communicate with the agent by text, files, files+message, and in the future have something like app.live_query(user_id...) that would open a websocket for bidi and handle everything, that would be huge.
Describe alternatives you've considered
The current alternative is to use a Runner and deploy to Cloud Run or anywhere else and do it manually. But I think it would be much better to have the ability to deploy directly to Agent Engine. It makes the agent development so much easier and smoother.