magick_mind.resources.v1.chat

V1 chat resource implementation.

ChatResourceV1

class ChatResourceV1(BaseResource)

Chat resource client for V1 API.

Provides typed interface for sending chat messages to mindspaces.

Example:

response = await client.v1.chat.send( api_key="sk-...", mindspace_id="mind-123", message="Hello!", enduser_id="user-456", fast_model_id="gpt-4", smart_model_ids=["gpt-4"], compute_power=50, ) print(response.content.content) # AI response text

ChatResourceV1.send

async def send(api_key: str,
               mindspace_id: str,
               message: str,
               enduser_id: str,
               fast_model_id: str = "",
               smart_model_ids: Optional[list[str]] = None,
               compute_power: int = 0,
               config: Optional[ConfigSchema] = None,
               reply_to_message_id: Optional[str] = None,
               additional_context: Optional[str] = None,
               artifact_ids: Optional[list[str]] = None) -> ChatSendResponse

Send a chat message to a mindspace and get AI response.

Arguments:

api_key - API key for LLM access
mindspace_id - Mindspace/chat conversation ID
message - User message text to send
enduser_id - End-user identifier
fast_model_id - Model ID for fast brain
smart_model_ids - Model IDs for smart brain
compute_power - Compute power setting (0-100), defaults to 0
config - Optional ConfigSchema override — if provided, used directly and the flat model params (fast_model_id, smart_model_ids, compute_power) are ignored. Useful for advanced callers who already have a ConfigSchema.
reply_to_message_id - Optional ID of message being replied to
additional_context - Optional additional context for the message
artifact_ids - Optional list of artifact IDs to attach to message

Returns:

ChatSendResponse with AI-generated response

Raises:

ValidationError - If message is empty, mindspace_id invalid, or required fields missing
ProblemDetailsException - If mindspace not found (404), permission denied (403), or server error (500+). Always includes request_id for support.
RateLimitError - If API rate limit exceeded (429)
AuthenticationError - If JWT token is invalid or expired (auto-refreshed transparently)

Example:

Basic chat with flat params (preferred)

response = await chat.send( api_key="sk-test", mindspace_id="mind-123", message="What's the weather?", enduser_id="user-456", fast_model_id="gpt-4", smart_model_ids=["gpt-4"], )

Chat with attached artifacts

response = await chat.send( api_key="sk-test", mindspace_id="mind-123", message="Analyze these documents", enduser_id="user-456", fast_model_id="gpt-4", smart_model_ids=["gpt-4"], compute_power=80, artifact_ids=["art-123", "art-456"], )

Advanced: pass a pre-built ConfigSchema directly

response = await chat.send( api_key="sk-test", mindspace_id="mind-123", message="Hello", enduser_id="user-456", config=ConfigSchema(fast_model_id="gpt-4", smart_model_ids=["gpt-4"]), )

print(f"AI - {response.content.content}") print(f"Message ID: {response.content.message_id}")

Chat (v1)