Skip to content
Set up your Agent

Reply

Send plain text, markdown, attachments, and interactive cards from agent handlers.

A reply sends a message back into the conversation. Agents can reply with plain text, markdown with files, or interactive cards, depending on provider capabilities.

Interactive cards include buttons, dropdowns, links, and text inputs. When a user interacts with a card, the onAction handler fires with the action ID and selected value.

Use replies when the agent needs to communicate something to the participant in the conversation.

Replies are user-facing messages your agent sends back into the conversation. Use signals when you need to update conversation state, trigger workflows, or resolve the thread without messaging the user.

The public API is ctx.reply(content, options?). You can also return a string or JSX card from a handler instead of calling ctx.reply() directly.

Reply types at a glance

The following table summarizes the reply types your agent can send:

TypeContentAttachmentsUser interaction
Plain textStringVia options.filesNone
MarkdownString with markdownVia options.filesNone
Interactive cardsCard and child componentsNot on cardsButtons, dropdowns, links, inputs trigger onAction

To change a message after you send it, see Edit sent messages.

Plain text

Send a simple string reply with ctx.reply():

await ctx.reply('Hello! How can I help?');

Markdown

Send formatted text by passing a markdown string to ctx.reply():

await ctx.reply('**Report generated.** See the attached PDF.');

Sending attachments

Include files with string or markdown replies via the optional second argument.

When sending attachments, keep these limits in mind:

  • Attachments are limited to 25 MB per file.
  • Files are only supported with string or markdown replies, not card replies.
  • Provide each file with exactly one of url or data.

File reference type

Each file uses a FileRef object with the following shape:

type FileRef = {
  filename: string;
  mimeType?: string;
  data?: string | Uint8Array | ArrayBuffer | Blob;
  url?: string;
};

Use url for larger files. Novu fetches public HTTP(S) URLs server-side. Use data for small generated files in memory.

The following examples show both approaches:

await ctx.reply('Here is your report.', {
  files: [{ filename: 'report.pdf', mimeType: 'application/pdf', url: reportUrl }],
});
 
const csv = new TextEncoder().encode('name,total\nNovu,42');
await ctx.reply('CSV generated.', {
  files: [{ filename: 'report.csv', mimeType: 'text/csv', data: csv }],
});

Interactive cards

Cards are structured messages with buttons, dropdowns, links, and more. Build them with function calls or JSX.

Function call API

The following example builds a card with the function call API:

import {
  Card, Button, CardText, Actions,
  Select, SelectOption, Divider, CardLink,
} from '@novu/framework';
 
await ctx.reply(Card({ title: 'Order #1234', children: [
  CardText('Your order is ready for pickup.'),
  Divider(),
  Actions([
    Button({ id: 'ack', label: 'Acknowledge' }),
    Button({ id: 'escalate', label: 'Escalate', style: 'danger' }),
  ]),
  CardLink({ url: 'https://example.com/order/1234', children: 'View details' }),
] }));

JSX API

Configure tsconfig.json with "jsxImportSource": "@novu/framework", then return JSX from a handler or pass it to ctx.reply(). For a full JSX card example with Card, CardText, Actions, and Button, see Build your first agent.

When a user clicks a button or selects a dropdown value, onAction fires with actionId and value. See Handle events.

Available card components

The following table lists the card components you can use in replies:

ComponentDescription
CardContainer with an optional title
CardTextText block inside a card
ButtonInteractive button; id maps to actionId in onAction
ActionsRequired wrapper around Button elements
Select / SelectOptionDropdown; triggers onAction with selected value
DividerVisual separator
CardLinkClickable link
TextInputText input field

On this page

Edit this page on GitHub