from typing import List
from notebridge import Bridge, ChatMessage, ChatContext, AgentResponse
import openai
class MyAgent(Bridge):
def on_receive(self,
message_stack: List[ChatMessage],
context: ChatContext,
storage: dict) -> AgentResponse:
# TODO: Implement your agent here.
# You can access environment variables by using something like `os.environ['OPENAI_API_KEY']`.
gpt_messages = [{
"role": "system",
"content": f'You are role-playing as a physician called NoteAid, who can answer patients\' questions about their health. Here is patient\'s clinical note: {context.note}',
}]
for prev_message in message_stack:
if prev_message.is_agent:
gpt_messages.append({
"role": "assistant",
"content": prev_message.content,
})
else:
gpt_messages.append({
"role": "user",
"content": prev_message.content,
})
chat_completion = openai.ChatCompletion.create(model='gpt-4', messages=gpt_messages)
answer = chat_completion.choices[0].message.content
# `messages` is a list of messages that you want to send back to the user.
# `storage` is a dictionary that you can use to store data between different requests.
# You need to pass the dict into here in order to access it in the next request.
return AgentResponse(messages=[answer], storage=storage)
