Skip to main content

Documentation Index

Fetch the complete documentation index at: https://aimnahai.mintlify.app/llms.txt

Use this file to discover all available pages before exploring further.

Create Chat Agent

Chat agents handle text-based conversations on your website via the embeddable widget or the API.

Step 1: Create the agent

  1. Go to AgentsCreate Agent
  2. Set Agent Mode to Chat or Both
  3. Write your system prompt
  4. Select an LLM model
  5. Save

Step 2: Configure chat settings

Navigate to the Chat settings tab on your agent:

Chat Settings

SettingDescriptionDefault
Welcome MessageFirst message shown when chat opens
Input PlaceholderPlaceholder text in the input field”Type a message…”
Response DelayArtificial delay before responding500ms
Typing IndicatorShow animated dots while agent thinksEnabled
Auto-Close MessageMessage shown when chat closes from inactivity
Inactivity TimeoutClose chat after this period of silence60 min

Widget Appearance

SettingDescriptionDefault
PositionBottom-right or bottom-leftBottom-right
Primary ColorButton and header color#0066FF
Header TextTitle in the chat window header”Chat with us”
Avatar URLImage shown in the header
Button SizeFloating button diameter60px
Window SizeChat window width and height380 x 600
Powered ByShow “Powered by Aimna” badgeEnabled

Step 3: Enable the widget

Toggle Chat Widget to enabled. This activates the public widget endpoints for your agent.

Step 4: Get the embed code

  1. Expand the Embed Code section (appears after enabling the widget)
  2. Choose Chat Widget or Callback Widget
  3. Copy the script snippet
  4. Paste on your website before </body>
See Website Widget for full embedding instructions.

Step 5: Add a knowledge base (optional)

Give your agent access to your documentation, FAQs, or product information:
  1. Go to the Knowledge Base tab
  2. Click Add Source
  3. Upload files (PDF, DOCX, TXT) or enter a website URL
  4. The content is automatically chunked, embedded, and stored for retrieval
Your agent will search the knowledge base for relevant context before responding to each message.