BrowserSkill connects any shell-capable AI agent to the user's real Chrome or Edge browser so it can read pages, fill forms, and debug sites with existing login state, while tasks run in a separate, visible Agent Window.
Project overview
It lets agents reuse the browser's current login sessions instead of separate accounts or credentials, and works with any shell-capable agent the user chooses, providing only the browser connection.
Project type
AI Agent
Use cases
Automation
Deployment
Refer to project documentation
License
MIT
Best for
Developers and AI engineers who already use a shell-capable AI agent and need it to operate their real, logged-in Chrome or Edge browser for reading pages, filling forms, or debugging websites.
Key capabilities
Inspect page text and controls, click and type, manage tabs, capture screenshots, and upload or download files in local mode, using the browser's current login state.
Tasks run in a separate, visible Agent Window; an existing tab can be explicitly borrowed and returned when the task ends.
Connect actions to requests, response bodies, console messages, and page changes; inspect performance, slow APIs, and suspected duplicate requests; use HTTP rules or request replay to test a hypothesis.
Name browser instances, bind a task to a specific profile, or select a browser explicitly per session.
Adds native browser_* tools, task previews, and screenshot results to the DeepSeek Harness Web UI using the same CLI and extension.
Limitations and risks
Nested scrolling panels and virtualized lists have limitations in full-page capture.
Other Chromium browsers may work, but compatibility is not guaranteed; the documented requirement is Chrome or Microsoft Edge based on Chromium 125 or later.
Remote mode currently does not support file upload or download; remote mode also involves an optional authenticated WSS connection to a server running the agent and daemon.
Getting started
Install the CLI via install.sh, install the extension from the Chrome Web Store or Edge Add-ons and enable local connection, run bsk install-skill then bsk doctor, and verify by asking the agent to open https://example.com and summarize the page. Difficulty reflects the CLI script install, extension store install, and the skill installation plus connection check.
Evidence and sources
GitHub project description: CLI + extension for browser automation across any shell-capable AI agent.
README: BrowserSkill connects your AI agent to Chrome or Microsoft Edge, using the accounts you are already signed into.
README: Use the browser's current login state to read documents, search internal sites, fill forms, and complete web workflows.
README: Extension | Chrome and Microsoft Edge, based on Chromium 125 or later. Other Chromium browsers may work; compatibility is not guaranteed.
README: curl -fsSL https://raw.githubusercontent.com/Tencent/BrowserSkill/main/install.sh | sh