hanzi-browse and hanzi-in-chrome
These are competing implementations of the same capability—both enable LLM agents to automate browser interactions—with hanzi-browse appearing more actively maintained and adopted (higher downloads and stars), making hanzi-in-chrome largely redundant.
About hanzi-browse
hanzili/hanzi-browse
let any ai agent use the local browser
Implements a Chrome extension + Node.js backend that bridges AI agents to real browsers via MCP (Model Context Protocol), enabling authenticated web interactions like form-filling and file downloads directly from the user's signed-in session. Offers dual deployment modes: managed cloud execution ($0.05/task) using Gemini, or local BYOM with user-provided API keys. Includes pre-built skills for common workflows (LinkedIn prospecting, accessibility auditing, social posting) and a REST API for embedding browser automation into products.
About hanzi-in-chrome
hanzili/hanzi-in-chrome
any llm browsing for you in chrome
Scores updated daily from GitHub, PyPI, and npm data. How scores work