API Output Component

Use the API Output component to structure and standardize the final output of a workflow. It's especially powerful when used as the final step in an Agent Skill, ensuring that any workflow acting as an API endpoint returns a clean, predictable, and well-formed JSON response.

Why this matters

Consistent and clean outputs are crucial for building reliable agents. The API Output component acts as a formal "return" statement for your workflows, making them easier to call, debug, and integrate with other systems.

What You’ll Configure

• Define Output Fields
• Select an Output Format
• Handle the Final Output
• Best Practices
• Troubleshooting Tips
• What to Try Next

Step 1: Define the Output Fields (Inputs)

The "inputs" of this component are actually the fields you want to include in your final JSON output object. For each key-value pair you want in the response, you will add one input.

Field	Required?	Description
Name	Yes	The key name for the field in the final JSON output (e.g., data).
Type	Yes	The data type (e.g., String, Number, Array, Object).
Optional	No	Marks the field as not required to be present in the output.
Default Value	No	A fallback value to use if the connected input is empty.

Structuring a Response

Think of each input as a field in your final JSON response. If you want a response like {"status": "success", "data": {...}}, you would create two inputs named status and data.

Step 2: Select an Output Format

This setting controls the structure and level of detail in the final JSON object.

Full (Default): Returns a detailed object including component metadata like names and descriptions for each field.

Ideal for: Debugging and getting a comprehensive view of the output structure.

Step 3: Handle the Final Output

The component produces a single, consolidated JSON object based on your defined inputs and selected format.

Output	Description	Data Structure
Output	The final, structured JSON object representing the workflow's result.	A valid JSON object. The exact structure depends on the Output Format.

Example (Minimal Format): If you have two inputs, status (value: "ok") and user_id (value: 123), the output would be:

{
  "status": "ok",
  "user_id": 123
}

Best Practices

• End Skills with API Output: Always use API Output as the final step for any Agent Skill that has "Advanced Request Parts" enabled. This creates a formal, reliable contract for your skill's response.
• Standardize Your Responses: Adopt a consistent response structure across all your skills (e.g., { "success": true, "data": {...} } or { "success": false, "error": {...} }). This makes your agent more robust and easier to integrate with.
• Use Minimal for Production: The Minimal format is almost always the right choice for live skills, as it provides a clean and predictable output without unnecessary metadata.
• Connect All Inputs: Ensure every input field in the API Output component is connected to the data it's supposed to represent from earlier in your workflow.

Troubleshooting Tips

If your output is incorrect...

• Check the Output Format: The most common issue is having the wrong format selected. If your output contains extra metadata you don't want, switch from Full to Minimal.
• Missing Fields: If a field is missing from your output, check that its corresponding input is correctly named and connected to a data source in your workflow.
• Incorrect Data Structure: Ensure the data you are passing into the inputs matches the Type you selected for each. For example, don't pass a string into an input that is typed as an Object.

What to Try Next

• Make this the final component in an Agent Skill to define its return value.
• Call a skill that uses API Output from an external application to see how it receives the clean, structured JSON response.