Skip to content

Agentic WorkFlow Platform Guide

Agentic WorkFlow lets you automate things you do on websites (copying info, clicking buttons, filling forms, extracting data, summarizing content with AI) using a visual, drag-and-drop builder.

You don’t need to be a developer.

Agentic WorkFlow runs entirely inside your browser. That means it’s fast, private, and can directly interact with the page you’re currently viewing.

Here’s the mental model:

1. You build a workflow Connect blocks (called “nodes”) to describe what should happen.

2. A trigger starts it For example: you run it manually, a page loads, you click a button, or you use the page context menu.

3. Nodes do the work Nodes can extract data from the page (text, links, tables), interact with the page (click, scroll, fill fields), or process data (format, filter, summarize with AI).

4. You review the result You can see outputs step-by-step and adjust your workflow until it matches what you want.

GoalDescriptionComplexity
Research HelperExtract text, summarize with AI, and save notesBeginner
Smart Form FillerAnalyze a form and fill it intelligentlyIntermediate
Content MonitorWatch sites for changes and notify youIntermediate
Research AgentPlan research, visit sites, and write a reportAdvanced
Multi-Agent SystemCoordinate multiple AI agents for complex tasksExpert
  • It runs in your browser (so it can work with the page you’re on).
  • You can automate both “reading” (extracting) and “doing” (clicking, typing, submitting).
  • You can use AI locally (private/offline) or connect a cloud AI provider.
  • Visual editor: build by dragging and connecting nodes
  • Node library: pick actions and triggers like building blocks
  • Test as you go: run one step or the full workflow
  • Safe iteration: adjust parameters and rerun until it works
  • Extract data: selected text, full page text, links, images, tables, metadata
  • Interact with the page: click, scroll, hover, fill fields, submit forms
  • Wait for conditions: e.g. “wait until this element appears”
  • Local models: run AI privately (when available)
  • Cloud models: connect to external AI services if you choose
  • Knowledge base: store information locally and search it (RAG)
  • AI reasoning: turn raw page content into useful output
  • Data Transformation: Clean, format, and structure data
  • API Integration: Connect to external services and databases
  • File Operations: Save, load, and process various file formats
  • Conditional Logic: Smart branching and decision making

Remember: if you can explain the steps (“get this text”, “click that button”, “summarize this page”), you can usually turn it into a workflow.