> ## Documentation Index
> Fetch the complete documentation index at: https://docs.4minds.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat controls

> Steer SYMI mid-response, queue messages, set reasoning effort, and work with to-do lists and questions.

## Steer or queue a message

You can keep typing while SYMI is responding. Choose what your message does.

| Option | What it does |
| - | - |
| **Queue** | SYMI finishes its current response, then answers your message as its next turn. |
| **Steer** | SYMI reads your message in the middle of its current response and adjusts, without starting a new turn. |

Queue is the default. Press **Enter** to use it. Press **Cmd+Enter** (Mac) or **Ctrl+Enter** (Windows and Linux) to use the other option.

To change your default, open **Settings** and set **Enter behavior while busy** to **Queue** or **Steer**.

> [!NOTE]
>
> Steering works for plain text messages. A message with an attachment is queued instead.

### Manage queued messages

Queued messages appear above the chat box, one per line. Select the X at the right of a message to remove it from the queue. With two or more queued, select the header, for example **2 queued**, to open or fold the list. You can queue up to 5 messages in a conversation. A message, including its attached files, can be up to 65,536 tokens (about 65K). See [Limits](https://docs.4minds.ai/symi/how-symi-works.md#limits).

### Stop a response

Select the stop button to end SYMI's current response, along with any helpers and background commands it started. Messages you queued are not discarded. The next one starts when the response stops. Remove a message from the queue first if you do not want it sent.

## Reasoning effort

Use the reasoning control in the chat box to choose how long SYMI thinks before it acts.

| Level | What it does |
| - | - |
| **High** | SYMI's usual thinking. This is the default. |
| **Max** | SYMI thinks much longer. Responses are slower, and a long reply may be cut off. |

Your choice is saved for the conversation you are in and applies from your next message. A new conversation starts with **High**. It controls SYMI's own reasoning, including when you work with a 4Minds model.

## To-do lists

For long, multi-step work, SYMI writes a to-do list in a **To-dos** strip above the chat box. Each item shows its status: completed, in progress, stopped, or pending.

* When SYMI finishes a response, any item still in progress is marked completed.
* When you stop a response, items that were in progress are marked **stopped**.
* The list clears when you send your next message.

SYMI skips the list for short, single-step requests.

## Questions before big tasks

Before a large task, such as a website, document or report, SYMI may ask a few questions that shape the result. A question card appears above the chat box.

* Select an option, or choose **Type your answer** to write your own.
* Select **Skip this question** to let SYMI use its best judgement. On the last question, select **Submit** to send your answers.

If you already gave the details, or tell SYMI to just do it, it does not ask. Scheduled jobs never ask, because nobody is there to answer.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.