BrowserSkill
SHA-256CLI + extension that lets AI agents like Cursor or Claude Code drive your real, logged-in Chrome or Edge browser. Unlike Playwright MCP, it reuses your existing sessions and keeps tasks in a separate window.
Smart Download
Download Download Version
vcli-v0.3.3 · 6.2 MB
Let AI agents work in your real, logged-in browser in a separate window without interrupting you.
Core Features
- Reuse your existing logged-in browser sessions, no re-login required
- Tasks run in a dedicated, visible Agent Window you can watch or take over
- Website debugging with request, response body, and console evidence
- Full-page screenshots, silent MP4 task recording, and table export to JSON/CSV
- One command installs the skill for Cursor, Claude Code, Codex, and more
What It Can't Do
- •Only Chrome and Edge (Chromium 125+) are officially supported; other Chromium browsers are not guaranteed to work. It is not a standalone app and requires a shell-capable AI agent—there is no GUI of its own. Automation uses your currently logged-in accounts, so mind authorization scope and privacy. Native page popups may briefly steal window focus, and zero interruption cannot be guaranteed. Only use the debugging features on websites you are authorized to test.
Use Cases
- Automate form filling and workflows on internal sites that require your login
- Debug website bugs by letting the agent capture network and console evidence
BrowserSkill is an MIT-licensed tool from Tencent that connects shell-capable AI agents (Cursor, Claude Code, Codex, DeepSeek Harness, and more) to your real Chrome or Edge browser via a `bsk` CLI daemon plus a browser extension. Its standout advantage over alternatives like Playwright MCP or browser-use is that it works with your existing logged-in accounts instead of a fresh sandboxed profile, eliminating re-login friction for internal sites and authenticated workflows. Tasks run in a dedicated, visible Agent Window so you can keep working, watch progress, or take over for human-only steps like captchas. Beyond basic clicking and typing, it offers website debugging with request/response/console evidence, full-page screenshots, silent MP4 task recording, structured table extraction to JSON/CSV, element waits, and multi-browser instance management. Compared with heavier headless automation frameworks, it is lighter to adopt: install a single binary, add the extension, run one skill-install command, and your agent gains browser superpowers with audit and history export for review.
Tags
Getting Started
Download installer
Click the button above to download the installer for your system
Install the software
Double-click the downloaded installer and follow the prompts
Run the official install script in your terminal to install the bsk CLI (macOS, Linux, or Windows)
Install the BrowserSkill extension from the Chrome Web Store or Edge Add-ons and enable the local connection
Run bsk install-skill to set up the skill for your AI agent
- Run the official install script in your terminal to install the bsk CLI (macOS, Linux, or Windows)
- Install the BrowserSkill extension from the Chrome Web Store or Edge Add-ons and enable the local connection
- Run bsk install-skill to set up the skill for your AI agent
SHA-256 checksum verified
Checksum extracted from GitHub official Release page
SHA256 Checksum
5ed6aa00e60c0538a637febac6bfbc6a0bd55a11083cd90a987968cff4bc0e02This checksum is extracted from the GitHub Release page. Verify file integrity after download.
All SHA-256 checksums on this platform are extracted from the project's official GitHub Release page, without any modification. You can independently verify them on the GitHub Releases page.
Open Source Transparency
View GitHub SourceUninstall Info
Delete the bsk binary (default ~/.local/bin/bsk) and its data directory (default ~/.browserskill), remove the BrowserSkill extension from your browser's extensions page, and delete the browser-skill folder from each agent's skills directory.
No Extra Dependencies
Ready to use after download. No additional runtime required.
Having issues? Check the FAQ below
5 FAQs
Similar Projects
gemini-cli
Google's open-source terminal AI agent for coding, debugging, and workflow automation powered by Gemini models.
ollama
Ollama lets you download, run, and manage large language models locally. One command, multiple platforms, endless possibilities.
ComfyUI
ComfyUI is a modular AI engine with a node-based graph interface for generating images, video, 3D models, and audio. Unlike traditional webui tools, it gives pixel-level control over every model and parameter, making it the preferred choice for professionals.