> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mossx.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Input box
> Use @, /, #, and ! completions to send files, images, and prompts.
The input box is where you give the model instructions. The placeholder lists the full set of features:
```text theme={null}
@ cite a file, # pick an agent, ! insert a prompt, Enter to send
```
## Send and newline
**Settings → General → Behavior** has two send styles:
| Option | Send | Newline |
| -------------------- | --------------------------- | ------------- |
| Enter to send | `Enter` | `Shift+Enter` |
| ⌘/Ctrl+Enter to send | `Cmd+Enter` or `Ctrl+Enter` | `Enter` |
While generating, the send button becomes stop. The input box grows with longer messages.
## Four completions
### `@` cite a file
Type `@` to search project files or directories. A selected item becomes a file chip. The model receives the path, not the full file contents.
You can also write a line-range citation:
```text theme={null}
@src/Main.kt#L14-L40
```
Citations sent from an editor selection or the project tree use the same format. See [Editor integration](/en/jetbrains/ide/editor).
### `/` slash commands
Type `/` to list commands from the current runtime. The plugin merges the SDK list with local commands.
These local commands are always available:
| Command | What it does |
| -------- | ---------------------------------------------- |
| `/clear` | Clear the current chat and start a new session |
| `/new` | Same as `/clear` |
| `/reset` | Same as `/clear` |
`/compact` comes from the Claude runtime and compresses context. Claude also supports `/mcp` to inspect MCP status.
Some commands work only with a specific provider. If you picked the wrong runtime, the UI says it is only supported by that provider.
### `#` pick an agent
Type `#` to choose a configured agent. After you select one, that agent’s system prompt is applied to later turns.
Create, import, or export agents in **Settings → Agents**. See [Agents and prompt library](/en/jetbrains/customize/agents-prompts).
### `!` insert a prompt
Type `!` to open the prompt library. Selecting an item inserts the template at the cursor. You can edit it before sending.
Maintain prompts in **Settings → Prompt library**.
## Attachments, paste, and drag-and-drop
The input box accepts:
* File paths dragged from the project tree
* Images dragged or pasted in
* Code citations from an editor selection
* Citations from selected console text
Images show a thumbnail. You can remove them before sending.
## Bottom selectors
From left to right under the input box you usually see:
1. **Config / process**: inspect the current daemon and restart it if needed
2. **Provider**: Claude Code, Codex, Grok, Kimi, OpenCode, Pi
3. **Model**: models for the current provider
4. **Mode**: default, plan, agent, auto
5. **Attach / stop / send**
Claude Code also has extra switches such as reasoning strength. Codex has Fast mode. CLI providers do not offer plan mode.
## Message queue
You can type the next message while the previous turn is still generating. Queued messages send automatically after the current turn ends.
## Prompt enhancer
The input toolbar can rewrite the current draft before you send it. Which provider and model do the rewrite is set in **Settings → Prompt enhancer**. See [Prompt enhancer](/en/jetbrains/customize/prompt-enhancer).
## History input completion
With **Settings → Other → History input completion** on, typing suggests earlier inputs. Press `Tab` to accept.