How to pass hardcoded dynamic messages as agent's responses in the chat history ?

[HG2407]

Please read this first

• Have you read the docs?Agents SDK docs
• Have you searched for related issues? Others may have had similar requests

Question

Suppose I have a todo tracking agent, which tracks the todo tasks of a user. Now on the first message I want to provide the list of todo tasks of the user with a properly phrased hardcoded message from my end. I also want the model to have context of these tasks. I can't add these todo tasks in the system prompt. What are my options ?

I tried doing this:

input_items.append({'role': 'assistant', 'id': "msg_MADE_UP_ID", 'content': [{'annotations': [], 'text': hardcoded_content, 'type': 'output_text'}], 'status': 'completed', 'type': 'message'})
input_items.append({"content": user_input, "role": "user"})
result = await Runner.run(current_agent, input_items, context=context)

But I get this error:

openai.NotFoundError: Error code: 404 - {'error': {'message': "Item with id 'msg_MADE_UP_ID' not found.", 'type': 'invalid_request_error', 'param': 'input', 'code': None}}

[HG2407]

I want to pass this hardcoded content as the first message of the chat. Is there a separate type: system similar to user and assistant ?

[rm-openai]

Just don't set the ID.

    agent = Agent(name="Assistant")

    inputs = [
        {"role": "user", "content": "What's your name?"},
        {"role": "assistant", "content": "My name is Jimmy the AI, how can I help you today?"},
        {"role": "user", "content": "What's your name again?"},
    ]

    result = await Runner.run(agent, inputs)

[rm-openai]

And yes, valid roles are "user", "assistant", "system", "developer"

[github-actions]

This issue is stale because it has been open for 7 days with no activity.

[github-actions]

This issue was closed because it has been inactive for 3 days since being marked as stale.