WebOperator is an open-source Chrome extension by the developer KazKozDev that takes a goal typed in plain English and carries it out in the browser tab you are looking at. It writes a plan you can read, calls one browser tool at a time, and checks each result against a fresh snapshot of the page before moving on. It can navigate, click, type, scroll, switch tabs, take screenshots and pull out structured text, so a single task can run across several open tabs. The side panel streams the plan, every action and the final answer, and keeps the full trace in a history view. Scheduled runs use Chrome alarms to repeat a task once, hourly, daily or weekly with the side panel closed, and a run that meets a login wall or a verification challenge stops and marks itself as needing the user. The same browser is also exposed to other agents as an MCP server. The weboperator-mcp bridge is published on npm and listed in the official MCP registry, and it offers ten tools over stdio, including browser_snapshot, browser_navigate, browser_click and weboperator_execute_goal, with ready-made configurations for the Hermes and OpenClaw agents. It is aimed at individual developers and technical users who want browser automation running locally inside their own signed-in Chrome, with the choice of model left to them.
WebOperator is free and MIT-licensed, with no paid tier and no hosted service. The extension is loaded unpacked from a GitHub release or from a local build rather than installed from the Chrome Web Store. The vendor lists Chrome and Brave on macOS and Linux as the tested platforms, with Windows untested and the MCP bridge installer limited to macOS and Linux. Out of the box it talks to a local Ollama model, so it can run with no API key at all. It also accepts keys for Anthropic, OpenAI, Gemini, DeepSeek, OpenRouter, xAI, SiliconFlow, MLX and any OpenAI-compatible endpoint, and usage on those providers is billed by the provider rather than by WebOperator. The one hard requirement of the model is tool calling. A remote provider sees the page text and screenshots the agent sends it, which the vendor states plainly in its own documentation.
WebOperator suits developers who want a browser agent running locally in their own Chrome with a model they choose. Teams wanting a Python library and a hosted cloud API for browser agents should look at Browser Use. Teams wanting hosted, computer-vision-driven workflow automation with a free tier should look at Skyvern. Teams wanting a no-code browser automation extension with cloud execution should look at Axiom.ai.
