Introduction
BrowserSkill connects shell-capable AI agents such as Claude Code, Cursor, Codex, and others to your already logged-in browser. Instead of spinning up isolated sessions, agents borrow a specific tab explicitly, return it when the task is done, and leave the rest of your browser untouched, so you can keep working while automation runs in a separate visible Agent Window.
Key Features
- Reuse real login state so agents work with sites you are already signed into, without separate test accounts
- Browser tasks run in a separate, visible Agent Window, keeping your own browser usable
- Works with any agent that can call a shell through the
bskCLI, with no lock-in to a model or framework - Built-in human-in-the-loop handoff for captchas, logins, and confirmation dialogs
Use Cases
- Automating web tasks on authenticated sites you already have open
- Letting coding agents drive real browser state during development and testing
- Web scraping, form filling, and end-to-end workflows that require human-only steps
Technical Highlights
- Local runtime composed of the
bskCLI/daemon plus a Chrome/Edge extension - Cross-platform support for macOS, Linux, and Windows, with a one-line installer
- Ships an installable skill that teaches agent harnesses how to use the
bskCLI