browser-act-cli

browser-act-cli is a browser automation tool for AI agents and other automated workflows that need to use the web, pass blocked pages, run multiple tasks safely, and return clean web data for reasoning. Its own materials describe it as a browser layer for agents and show it being used to scrape Amazon bestseller listings and export the results to CSV.

The tool’s listed capabilities include stealth fingerprints, TLS rotation, and residential proxies to help avoid web blocks. It also offers automatic CAPTCHA handling, with support named for reCAPTCHA, Cloudflare Turnstile, DataDome, and HUMAN Security. For harder stops such as two-factor authentication, it provides remote-assist so a person can take over from a mobile device or desktop and then return control to the agent. The documentation also mentions browser modes for reusing local Chrome login state, using stealth private mode for bulk scraping, and using stealth fixed-identity mode for multi-account scenarios. Clean text output, indexed interactions, and semantic memory are presented as features designed for agent reasoning.

browser-act-cli runs on the user’s local machine. The materials say it can run multiple agents and tasks in parallel without account mixups or state pollution, because each browser has its own identity and each session has its own workspace. It is presented for agents using the web, and the examples show it connected to Claude Code in a demo workflow. The site also says no registration is needed for a copy-to-agent flow.

A 7-day free trial is advertised, with trial credits available with any subscription plan. The page also lists partners including AWS, Microsoft Azure, Google Cloud, Oracle, Alibaba Cloud, Huawei Cloud, BytePlus, and Baidu AI Cloud. The product is a CLI tool, and the materials identify a browser-act skill with a version shown as browser-act@1.3 in the demo.