On This Page
Introduction
The chat input (compose box) is where you write and send messages to agents, workflows, or your workspace. It is designed to be simple, fast, and flexible while keeping advanced options within reach.
Card Title
### What you’ll see
- A message area where you can type your prompt.
- A Send button to submit your message.
- A toolbar with quick actions:
- Attach files
- Slash Commands (Coming Soon)
- Agent Picker (Coming Soon)
- Text Size
- Model Selector
- Voice Input (speech-to-text)
- A message area where you can type your prompt.
- A Send button to submit your message.
- A toolbar with quick actions:
- Attach files
- Slash Commands (Coming Soon)
- Agent Picker (Coming Soon)
- Text Size
- Model Selector
- Voice Input (speech-to-text)
Coming Soon
javascript
function example() {\n console.log("Hello World");\n}Step 1
First step description
Step 2
Second step description
This is the introductory paragraph that describes what this section is about...
Notice
Your notification content goes here...
Tip:
Your helpful tip goes here...
Card Title New
Chat Input (Compose Box)
The chat input is where you write and send messages to agents, workflows, or your workspace. It is designed to be simple, fast, and flexible while keeping advanced options within reach.
What You'll See
- A message area where you can type your prompt
- A Send button (paper plane icon) to submit your message
- A toolbar with quick actions:
- Attach files
- Slash Commands Coming Soon
- Agent Picker Coming Soon
- Text Size
- Model Selector
- Voice Input (speech-to-text)
Writing Your Message
Click the message area and start typing. The box expands as you type (up to about half your screen height on small displays; on larger screens it caps around typical desktop heights).
- Use clear, specific language. If you need structured output (e.g., bullet points, steps, or a table), say so in your prompt.
- If you're discussing a document, consider attaching it first (see "Attach files" below) so the system can reference it.
Sending Your Message
Click the Send button (paper plane icon) to submit your prompt to the workspace/agent.
After sending, you'll see the system begin processing. Depending on tools used (e.g., web search, document retrieval), it may take a moment.
Toolbar Options
Attach Files
- Click the paperclip icon to attach files to your message
- Use this when you want the system to reference specific content (e.g., PDFs, images, or text)
- Tip: Add a note in your message about what to extract or focus on in the attachment
- Your workspace or admin defines which file types are supported
Slash Commands Coming Soon
- A quick-commands menu for inserting predefined actions or templates
- Marked "Coming Soon" and currently disabled. It will become interactive when released
Agent Picker Coming Soon
- A selector for choosing specific agents directly from the compose box
- Marked "Coming Soon" and currently disabled. It will become interactive when released
Text Size
- Adjusts the message box text size for better readability while you type
- Use this if you prefer larger or smaller text when composing long prompts
Model & Voice Controls
Model Selector
- Shows the currently selected AI model for this conversation (e.g., "gemini-2.5-flash-lite")
- Click to choose a different available model if your workspace supports multiple providers
Tip: Faster models are great for brainstorming; larger models are better for complex reasoning or longer outputs.
Voice Input (Speech-to-Text)
- Click the microphone icon to dictate your prompt by voice (where supported)
- Useful for quick notes, stream-of-consciousness drafting, or when typing isn't convenient
Tips for Better Results
- Be explicit about the format you want (summary, step-by-step, table)
- Provide context: what you already know, what you need next, any constraints (time, tone, sources)
- For long tasks, break your request into smaller parts and iterate
- When attaching files, mention their filename and what to extract
Troubleshooting
- Can't send? Ensure you're connected to the workspace and logged in
- Nothing happens after sending? Wait a few seconds; some tools (like web search) can take time
- Very long prompts: The box will stop growing after a point to keep the page usable; you can still scroll within it
- "Coming Soon" items are visible but inactive until the feature is released by your admin/team
Privacy & Safety
- Only attach files you're comfortable sharing with this workspace
- If your organization uses logging or auditing, your messages and attachments may be recorded for quality and compliance
1. Introduction & Purpose
The chat input (compose box) is where you write and send messages to agents, workflows, or your workspace.
It is designed to be simple, fast, and flexible while keeping advanced options within reach.
2. What you’ll see
1. A message area – where you can type your prompt.
2. Send button – to submit your message.
3. Attach files – to add files context.
4. Text size – to change font size in chat.
5. Model selector – to select your preffered LLM Model.
6. Voice input – to speak your input message inside the text field.
3. Writing your message
Click the message area and start typing. The box expands as you type (up to about half your screen height on small displays; on larger screens it caps around typical desktop heights).
Use clear, specific language. If you need structured output (e.g., bullet points, steps, or a table), say so in your prompt.
If you’re discussing a document, consider attaching it first (see “Attach files” below) so the system can reference it.
4. Sending your message
Click the Send button (paper plane icon) to submit your prompt to the workspace/agent.
After sending, you’ll see the system begin processing.
Depending on tools used (e.g., web search, document retrieval), it may take a moment.
5. Attach files
Click the paperclip icon to attach files to your message.
– Use this when you want the system to reference specific content (e.g., PDFs, images, or text).
– Tip: Add a note in your message about what to extract or focus on in the attachment.
– Your workspace or admin defines which file types are supported.
6. Text Size
– Adjusts the message box text size for better readability while you type.
– Use this if you prefer larger or smaller text when composing long prompts.
7. Model Selector
– Shows the currently selected AI model for this conversation (e.g., “gemini-2.5-flash-lite”).
– Click to choose a different available model if your workspace supports multiple providers.
– Tip: Faster models are great for brainstorming; larger models are better for complex reasoning or longer outputs.
8. Voice Input (speech-to-text)
– Click the microphone icon to dictate your prompt by voice (where supported).
– Useful for quick notes, stream-of-consciousness drafting, or when typing isn’t convenient.
Tips for better results
– Be explicit about the format you want (summary, step-by-step, table).
– Provide context: what you already know, what you need next, any constraints (time, tone, sources).
– For long tasks, break your request into smaller parts and iterate.
– When attaching files, mention their filename and what to extract.
Troubleshooting
– Can’t send? Ensure you’re connected to the workspace and logged in.
– Nothing happens after sending? Wait a few seconds; some tools (like web search) can take time.
– Very long prompts: The box will stop growing after a point to keep the page usable; you can still scroll within it.
– “Coming Soon” items are visible but inactive until the feature is released by your admin/team.