AI Agent node modules
The list below is a reference of every AI Agent node module the integration supports. You will be able to pick any of them on the workflow canvas while building automations.
Browse every AI Agent node module below — each one opens to show its input parameters, outputs, and connection requirements.
AI AgentAction
Input fields[3]
| Field | Type |
|---|---|
| System prompt | text |
User messageRequired | text |
| Response format | choice |
| Let AI decide which tools to call | boolean |
Output fields[3]
| Field | Type |
|---|---|
| Text | text |
| Usage | object |
| Prompt tokens | number |
| Completion tokens | number |
| Cache read tokens | number |
| Cache write tokens | number |
| Tool calls | array |
| ID | text |
| Name | text |
| AI credits used | number |
Templates
Look for AI Agent templates in the WEXTL template library.
Browse templates →Explore alternative apps
- Condition[2]Branch the workflow on a boolean rule — route matching data down the true path and everything else down the fallback (else) route.
- ContinueIn an error-handler branch, swallow the error and resume the workflow on the success path.
- Human in the LoopPause the run for a person to review, approve, or reject before the workflow proceeds.
- FunctionCall a built-in or your own custom function as a single workflow step, with typed inputs and a result output.
- JS CodeRun custom JavaScript in a secure sandbox — transform data, call helpers, and return a result to the next step.
References
- WEXTL® AI Agent Source Code connector implementation.
- Our users use cases and automations.
- Official API Documentation for AI Agent.
- WEXTL® visual workflow builder data.
Last updated: Jul 18, 2026 19:12