Tencent BrowserSkill
Lets AI agents drive your real, logged-in browser without interrupting you.
About
Tencent's BrowserSkill is an agent skill for browser automation with a twist: instead of spinning up a headless browser, it lets AI agents use your real, logged-in browser — sessions, cookies, and auth intact — without interrupting your work. It ships as a CLI plus browser extension that works across any shell-capable AI agent (Claude Code, Cursor, Codex, and friends), and it was one of the fastest-trending agent-skill repos in September 2026 (~3.9k stars, +1,350 in a day on the trending charts).
The convenience is also the risk surface: an agent driving your authenticated browser can reach anything you can, so sandboxing and least-privilege thinking apply. As a young project the docs and edge-case handling are thinner than mature automation stacks like Playwright, and long-running sessions may need babysitting.
What it can do
- Drive your real logged-in browser sessions
- CLI + extension usable from any agent shell
- No interruption of your own browsing
- Works with Claude Code, Cursor, Codex
- Session, cookie, and auth preservation
Setup
Install the CLI and browser extension from the Tencent/BrowserSkill repo; point any shell-capable agent at it to start automating your live browser.
Prerequisites
A Chromium-based browser and a shell-capable AI agent (Claude Code, Cursor, Codex, etc.).
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 58; Trending score 74/100 from researcher assessment.
★ 3,940 GitHub stars
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.