Skip to content

Let your other systems put the assistant to work

The editor of the autonomous agent Overdue invoice follow-up, with its widget, the highlighted Response mode set to Async (202), the Active switch on and a prompt that uses payload variables from the webhook body.The editor of the autonomous agent Overdue invoice follow-up, with its widget, the highlighted Response mode set to Async (202), the Active switch on and a prompt that uses payload variables from the webhook body.

Some work for your assistant does not start with someone typing. An invoice becomes overdue in your billing system, a payment fails, a weekly report is due. Autonomous agents let that other system start an AI agent with one webhook call: the agent reads the call’s data, works with the same tools and instructions as one of your widgets, and hands the result back.

  • Triggered by a webhook. Each agent has its own address. Your billing system, a script or an automation tool calls it with a JSON body.
  • A prompt that reads the call. Paste an example body and insert its fields as {{payload.…}} variables, so every call fills in the invoice number, the customer or whatever your system sends.
  • Runs for a real customer. Each call carries a personal access token that belongs to one of your customers, so every run is recorded against that customer and uses their saved connections.
  • Wait, or collect later. A synchronous agent answers in the same call. An asynchronous one returns a run id at once; your system polls the status address or receives the result at its own callback address.
  • Every run on record. The agent lists its recent runs with status, customer and output, and each run is a job you can follow step by step under AnalyticsAgent jobs.
The editor of the autonomous agent "Overdue invoice follow-up", with "Back to agents" and the red "Delete agent" button above it. The form shows Name "Overdue invoice follow-up", Widget "Demo Ltd Support chat", the highlighted "Response mode" select set to "Async (202)", the "Active" switch (on) with "Inactive agents reject webhook calls with 403.", and the Prompt with the variables {{payload.invoiceNumber}}, {{payload.customerName}} and {{payload.daysOverdue}}. Async answers the webhook at once and runs in the background; Synchronous (200) waits for the result.The editor of the autonomous agent "Overdue invoice follow-up", with "Back to agents" and the red "Delete agent" button above it. The form shows Name "Overdue invoice follow-up", Widget "Demo Ltd Support chat", the highlighted "Response mode" select set to "Async (202)", the "Active" switch (on) with "Inactive agents reject webhook calls with 403.", and the Prompt with the variables {{payload.invoiceNumber}}, {{payload.customerName}} and {{payload.daysOverdue}}. Async answers the webhook at once and runs in the background; Synchronous (200) waits for the result.
An autonomous agent pairs a widget's tools with a prompt filled from the webhook body.
What this screenshot shows
Screen
Agents (/admin/configure/autonomous-agents)
Highlighted
combobox “”, form-top-right, Async (202). chooses whether the webhook waits for the answer
Also visible
Name, Widget Demo Ltd Support chat, Active on; Prompt with {{payload.*}} variables
Purpose
hero of the autonomous agents release note: the agent editor

Events that need judgement no longer wait for a person to open a chat and paste in the details. The agent picks them up the moment they happen, uses the same tools and guardrails as your widget, and leaves a full record of what it did and for whom.

  1. Open ConfigureAutonomous Agents, select New, pick the widget whose tools the agent uses and the Response mode, and write the prompt.
  2. Signed in as a company admin, create an access token for one of your customers with their PIN or password, using the request shown in the agent’s webhook panel.
  3. Call the agent’s address with that token and a JSON body, and read the answer under View output.

The details are in Setting it up in 7 steps.

For every workspace. Creating and changing agents needs the Admin role. Every access token belongs to one active customer with a PIN or password, and creating one needs a company admin’s sign-in; the Access tokens tab is for reviewing and revoking tokens. Runs use the AI like any other conversation and count towards your usage and budgets.