Creating a custom action

Connecting an AI agent to one of your systems. Development work, and the design decisions matter more than the code.

Connecting an AI agent to one of your systems. Development work, and the design decisions matter more than the code.

Create one

1. Open the AI agent's actions.
2. Add an action and describe what it does.
3. Set the connection: the endpoint and how it authenticates.
4. Define what goes in and what comes back.
5. Say what happens when the call fails.
6. Test it, including the failures.

Describe it for the agent, not for you

The description is how the agent decides when to use it. "Looks up the status and expected delivery date of an order by order number" is usable. "Order API v2" is not.

Ask for the minimum

Every input is something the customer must supply before anything happens. An action needing three pieces of information will often not complete, because people abandon at the second question.

Return what the reply needs

Not the whole record. Sending back forty fields when the reply uses three makes the answer less predictable and puts data into the conversation that did not need to be there.

Failures are the design

System unavailable, nothing found, ambiguous match, permission denied. Each needs a defined outcome, and "hand over to a person" is a perfectly good one.

An action without failure handling produces conversations that stop mid-sentence.

Credentials belong in the connection

Never in a description, a reply, or a procedure. Use a service account scoped to exactly what the action does, so a mistake cannot reach further than intended.

Test before it exists in a conversation

Verify the call on its own first. Debugging an action inside a live exchange means two systems and a language model between you and the fault.

See also

• Testing actions
• Managing actions
• What actions are and what an AI agent can do with them