Skip to main content

Prompts, Context, and Messages

Dhenara supports multiple input styles.

1) Simple: prompt​

Use this for single-shot calls.

response = client.generate(prompt="Explain Kafka in one paragraph")
print(response.chat_response.text())

Use messages for chatbots and workflows. It’s provider-friendly and preserves the correct message structure (including tool calls).

from dhenara.ai.types.genai.dhenara.request import MessageItem, Prompt

messages: list[MessageItem] = [
    Prompt.with_text("Give me 3 ideas for a CLI tool"),
]

resp1 = client.generate(messages=messages)
messages.append(resp1.chat_response.to_message_item())

messages.append(Prompt.with_text("Now pick the best one and write a README outline"))
resp2 = client.generate(messages=messages)

print(resp2.chat_response.text())

3) Legacy style: context​

If you already have history in a separate list, you can pass it via context. For new code, prefer the Messages API.

Do you still need PromptFormatter?​

Usually no. Dhenara’s request/response types are designed so that multi-turn history can be passed back directly.

If you’re migrating older code that manually transforms history between providers, PromptFormatter can help, but we recommend starting from messages instead.