The chat input
A guided tour of every control in the chat input, explaining what each button does and why.
Last updated on
This is the chat input. Click any control to jump to what it does.
Click any control to jump to what it does.
Writing and sending
Type your question in the message box and send it. The assistant streams back an explanation and SQL.
| Key | Action |
|---|---|
| Enter | Send the message |
| Shift + Enter | New line |
The send button (bottom right) sends your message. While a reply is generating it becomes a Stop button; see Reading responses.
A typical answer comes back as a titled, ready-to-run query:
Attachments
The paperclip adds files for the assistant to read, handy for a screenshot of a report or a PDF spec. You can also paste an image or drag files onto the input. Accepted types and limits:
| Max total upload size | 100.0 MB |
| Max images per message | 20 |
| Max size per image | 20.0 MB |
| Max PDFs per message | 5 |
| Max size per PDF | 32.0 MB |
| Max pages per PDF | 100 |
| Accepted image formats | JPEG, PNG, GIF, WEBP |
Quick AI settings
The AI Settings button slides out a panel for quick tweaks, applied to your account: custom instructions (up to 1,000 characters), SQL comments, table aliases, brief explanations, report summary, and SQL column detail. It is a shortcut to the most-used AI settings without leaving the chat; "Advanced Settings" opens the full page.
Prompt type
The Reporting / Custom Field toggle tunes how your question is interpreted. Reporting (the default) is for normal report queries; Custom Field is for building HaloPSA custom-field dropdown SQL. Most questions stay on Reporting; see Reporting vs Custom Field for the detail.
New chat
New chat starts a fresh conversation with its own token budget. Start one for a new topic rather than letting a thread grow; it keeps answers focused and uses fewer tokens. Your previous chats stay in the history sidebar, so starting fresh never loses anything.
Token limit
The counter (e.g. 12.5k / 120k) shows how much of this conversation's budget
you've used; the bar along the top fills as you go. Click it to raise the limit:
pick a preset or a custom value, choose This chat only or Change
globally, tick the acknowledgement, and confirm.
You're charged in full for new messages; re-reading the conversation so far is cached, so it costs only a small fraction. Here's roughly how usage grows:
How a conversation grows
Drag to see roughly how many tokens a conversation uses as it goes on. You pay full price for each new message; re-reading the conversation so far is cached, so only a small fraction counts.
Message turns
10
Est. tokens used
50,343
input + output
Of the 120k default
42%
~23 turns fit
Estimates for a typical text conversation. Large file attachments use a lot more per message, so they reach the limit sooner.
For the full picture, see Tokens in chat and Understanding tokens.