# What is Open Comet?

Open Comet is an autonomous AI browser agent built as a Chrome extension. It uses a state-of-the-art "visual grounding" system to perceive web pages and execute multi-step tasks by reading, clicking, typing, and scrolling.

## Truly Autonomous

Unlike standard chatbots, Open Comet operates directly within your tab. It observes the visual state, reasons about the next step, and acts through native browser APIs.

## Key capabilities

### Autonomous Agent

Navigates and interacts with any page using screenshot + DOM analysis.

### High-Fidelity Scraper

Site-agnostic extraction of tables and structured data to JSON/CSV.

### Token Tracking

Real-time monitoring of AI consumption and costs across all models.

### Deep Research

Multi-source synthesis that queries, reads, and reports across many tabs.

### Skill System

Create reusable agent macros for complex, recurring workflows.

### Local AI (Ollama)

100% private and free agent stack running on your own hardware.

## How it works

01

**Input Task**

Describe your goal in natural language (e.g., "Find the cheapest flight").

02

**Perception**

The agent securely analyzes the active tab using Intelligent Payload Management.

03

**Reasoning**

The AI determines the next action using a self-healing intent resolver.

04

**Validation**

The runtime verifies the outcome of the action before proceeding.