Apify Agent [Preview]

Agent for driving browser actions. Record custom automation scripts or replay existing scripts in various formats.

Convert LLM-based agentic actions into stable and reliable code (future).

Part of Project Codify

Project Codify

Complete end-to-end browser automation pipeline running inside browser (or locally). Codify and replay browser actions as robust (deterministic) and reusable scripts. From rapid prototyping and quick automation scripts to sophisticated deployments at scale.

• Login - reusable authentication sessions - reusable secure login (TBD) ☑️

• Agent - turn scripts into browser actions - code or text script (future) ✅

• Coder - turn browser actions into scripts (TBD - currently integrated) ✅

• Robot - automation engine for scalability (TBD - currently integrated) ✅

• MCP - automations become reusable tools you can reuse through AI ✅

More is on the way... Give it a shot 🎬 or join the list to follow the project! 🔔

Why?

Problem:

• Browser agents are useful but not always reliable.
• Code is robust but coding from scratch is tedious.

Why not have the best from both worlds? 💯

Agent

Agent is the performer of browser actions to be replayed or recorded into reusable automation scripts. It manages automation scripts, from detection and normalization to execution and recovery.

Usage

Run the Agent actor. You will be guided through the entire process.

💡 Remove the default input script to start the recording of your own automation!

Overview

Agent module is designed to transform browser interactions into reliable, reproducible, and scalable automations. The source of the browser actions can be manual recording or agentic in nature*. It handles the execution flow, ensuring that scripts are validated, formatted correctly, and ready for execution.

Features

• input parameter detection
• automatic detailed logging
• automatic output and debugging tools
• automatic screen resolution detection
• code hints for script in the web editor
• support for different script formats (partial)
• automatic bot detection evasion for the replay
• automatic export to Apify task
• automatic export to AI MCP tool
• automatic export to script file

Highlights

• Multi-Format Support: Detects script formats and handles them appropriately. (Playwright, Puppeteer*, or native)
• Autonomous Normalization: Transforms raw web recordings and different script formats on the fly.
• Intelligent Flow Control: Manages the lifecycle of an automation task, including launching the Coder module to generate new scripts.
• Advanced Validation: Ensures that scripts are validated before execution.
• Smart Error Recovery: Implements strategies for code generation recovery.

How it Works

1. Script Ingestion: The Agent receives a script (from input or external source).
2. Detection & Validation: It analyzes the script and validates its integrity.
3. Transformation: If necessary, it transforms the script to a normalized format.
4. Execution Management: It hands off the processed script for high-scale execution.
5. Integration: If a script is missing it will trigger the Coder mode to generate it.

Agent turns reproducible, deterministic scripts into browser actions.

* future support