For some workflows, you might want to refine the automation graph or input variables the builder agent has generated. For that we offer a UI.

​
Edit Graph

The main steps of your business process are captured in the automation graph. A graph consists of nodes and edges. Nodes are predefined actions that you instruct our agent to take such as clicking, typing or making a decision. You can add nodes by right-clicking on the canvas and connect them with each other by drawing an edge between them.

​
Node Types

Action Nodes
  • Navigate: Opens a specified URL
  • Click: Clicks on an element using selectors or LLM-guided targeting
  • InputText: Types text into form fields with options for clearing and submitting
  • InputSelect: Chooses options from dropdown menus with fuzzy matching support
  • ExtractDatamodel: Pulls structured data from pages using JSON schemas
  • Scroll: Scrolls up or down the page
  • Screenshot: Captures a full-page screenshot with configurable scroll depth
Control Flow Nodes
  • BoolCondition: Creates branching logic with optional comparison operators (EQUAL, CONTAINS, etc.)
  • Loop: Executes repeated operations over arrays or numeric ranges
Specialized Nodes
  • Tfa: Handles two-factor authentication (SMS, email, authenticator apps)
  • FileUpload: Manages file uploads on websites using dowloadable file URLs as input variables
  • FileDownload: Downloads files with optional metadata
  • TabManagement: Controls browser tabs (open, close, switch)
  • UserInteraction: Pauses agent execution for manual user input or confirmation

​
Execution Types

Nodes can use different execution strategies:
  • STATIC: Uses explicit XPATH selectors for deterministic targeting
  • LLM_VISION: Uses AI to do an action or make a decision based off a screenshot.
  • LLM_DOM: Leverages AI to extract elements in the DOM structure (for ExtractDatamodel only)

​
Edges and Flow Control

Edges define how your workflow moves between nodes:
  • Sequential flow: Edges for straightforward progression
  • Conditional branching: true/false edges based on boolean conditions
  • Loop control: loop_done/loop_not_done for iteration management

​
Node Design

  • Use descriptive names for easy workflow understanding
  • Leverage LLM execution types when static selectors get too complex
  • Use STATIC execution when possible for speed and reliability

​
Edit Inputs & Outputs

​
Input Schema

Define the structure of data your workflow expects from users or systems. This is particularly useful for form-filling automations that need dynamic content and ensuring input data meets website validation requirements. You can use our visual editor or paste raw JSON schema.

​
Output Schema

Specify the structured data format your workflow returns after execution. This guarantess that the output will adhere to your defined schema.