> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/SmartEatAI/smart-eat-ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Get Conversation History

<Warning>
  This endpoint is currently not implemented in the backend. The chat system uses LangGraph's built-in memory management with thread-based context.
</Warning>

## Planned Implementation

Conversation history is maintained automatically by the AI assistant using thread IDs formatted as `{user_id}-{date}`. Each day creates a new conversation thread, and context is preserved within the same day's interactions.

## Alternative: Thread-Based Context

The SmartEat AI chat system currently uses LangGraph's state management to maintain conversation context:

* **Thread ID Format**: `{user_id}-{current_date}`
* **Context Window**: Conversations are scoped to the current day
* **State Persistence**: Message history is maintained in-memory during the session
* **Profile & Plan Access**: The assistant always has access to your current profile and active meal plan

## Example Current Behavior

When you send multiple messages in a day, the assistant maintains context:

```bash First Message theme={null}
curl -X POST https://api.smarteat.ai/api/chat/ \
  -H "Authorization: Bearer <token>" \
  -d 'message=Create a meal plan for me'

# Response: "I've created a 7-day meal plan..."
```

```bash Follow-up Message (Same Day) theme={null}
curl -X POST https://api.smarteat.ai/api/chat/ \
  -H "Authorization: Bearer <token>" \
  -d 'message=Can you swap the dinner for day 1?'

# Response: "I'll swap the dinner recipe from your meal plan..." 
# (Assistant remembers the plan it just created)
```

## Schema Definition

When this endpoint is implemented, it will return conversation history using the following schema:

### Response Structure

<ResponseField name="history" type="array">
  Array of messages in chronological order

  <Expandable title="Message object">
    <ResponseField name="role" type="enum">
      Role of the message sender

      * `user` - Message from the user
      * `chef` - Message from the AI assistant
    </ResponseField>

    <ResponseField name="content" type="string">
      The message content
    </ResponseField>

    <ResponseField name="timestamp" type="string">
      ISO 8601 timestamp of when the message was sent
    </ResponseField>

    <ResponseField name="original_recipe" type="object" nullable>
      Only present for suggestion messages - the original recipe being replaced
    </ResponseField>

    <ResponseField name="suggested_recipe" type="object" nullable>
      Only present for suggestion messages - the new recipe being suggested
    </ResponseField>

    <ResponseField name="accepted" type="boolean">
      Only present for suggestion messages - whether the user accepted the suggestion
    </ResponseField>
  </Expandable>
</ResponseField>

## Future Implementation Example

```json 200 - Success (Planned) theme={null}
{
  "history": [
    {
      "role": "user",
      "content": "Create a meal plan for this week",
      "timestamp": "2026-03-05T10:30:00Z"
    },
    {
      "role": "chef",
      "content": "I've created a 7-day meal plan for you with 3 meals per day...",
      "timestamp": "2026-03-05T10:30:05Z"
    },
    {
      "role": "user",
      "content": "Can you swap the dinner for day 1?",
      "timestamp": "2026-03-05T10:32:00Z"
    },
    {
      "role": "chef",
      "content": "I've found a similar recipe that matches your preferences...",
      "timestamp": "2026-03-05T10:32:03Z",
      "original_recipe": {
        "id": 1,
        "name": "Grilled Chicken Salad",
        "calories": 450
      },
      "suggested_recipe": {
        "id": 2,
        "name": "Mediterranean Chicken Bowl",
        "calories": 480
      },
      "accepted": true
    }
  ]
}
```

## Notes

* Conversation history may be implemented in a future version
* Current implementation relies on LangGraph's state management
* Each day starts a new conversation thread
* Consider implementing persistent storage for long-term conversation history