API Call

The API Call action allows your AI assistant to interact with external systems, services, or databases by sending HTTP requests. It supports all major HTTP methods and allows for custom pre/post-processing logic in Python.

This action is ideal for dynamic integrations like sending data, retrieving external information, triggering backend processes, or interacting with third-party APIs.

---

What It Does

When triggered, this action performs a direct HTTP request to the specified API endpoint. You can define the method, payload, and optionally include pre-processing and post-processing logic to manipulate the request or handle the response.

Use this for real-time data exchange or logic branching based on external API responses.

---

Action Interface

---

Configuration Options

API Endpoint (required)

Type: string The complete URL of the API endpoint to call.

Example: https://api.example.com/users

Method Type (required)

Type: string HTTP method used for the request. Choose from:

• GET
• POST
• PUT
• DELETE

Default: GET

Headers

Type: object
Key-value pairs sent with the request to provide metadata such as authentication, content type, or API configuration.

Example:

{
  "Authorization": "Bearer YOUR_TOKEN",
  "Content-Type": "application/json"
}

Payload

Type: object Key-value pairs to include in the request body or query params.

Example:

{
  "email": "@userEmail",
  "status": "active"
}

Pre Process Function

Type: python Python function that runs before the API call. Useful for preparing parameters or transforming data.

Example:

def preprocess(params):
    return {"email": params["email"].lower()}

Post Process Function

Type: python Python function that runs after receiving the API response. Use this to extract data or format it for the next step.

Example:

def handle(response):
    return response.get("data", {})

Only if everything was successful

Type: boolean When checked, the next action will only execute if this API call completes successfully.

---

Tips

• Use @parameter in the endpoint, payload, or code for dynamic inputs.
• Chain this with message generation or conditional actions based on response data.
• Return JSON or structured data to power follow-up tools like decision branches or summaries.

Let us know if you’d like ready-to-use API templates!