Repolode
投稿する
Tencent

Tencent/BrowserSkill

CLI and browser extension that let AI agents work in your logged-in Chrome or Edge while you keep using your browser.

GitHub で見る
スター
9,200
フォーク
661
未解決の Issue
105
最終コミット
4 日前

これは何？

BrowserSkill connects a shell-capable AI agent to Chrome or Microsoft Edge, using the accounts you are already signed into. Agent tasks run in a separate, visible Agent Window, and an existing tab can be explicitly borrowed and returned when the task ends. Agents can read pages, click and type, manage tabs, capture screenshots, and upload or download files in local mode.

The project is built around a `bsk` CLI with a background daemon and a browser extension. It is designed to work with agents such as Cursor, Claude Code, Codex, and others that can run shell commands, and it includes a dedicated plugin for DeepSeek Harness with native browser tools. Users choose the agent and model; BrowserSkill supplies the browser connection.

Beyond basic automation, it captures requests, response bodies, console messages, and performance data for website debugging, and can export loaded tables as JSON or CSV. A remote mode lets the agent run on a server while the browser and its logins stay on your computer.

主な機能

  • 01Runs agent tasks in a separate visible window that shares your browser login state.
  • 02Lets agents borrow an existing tab and return it when the task ends.
  • 03Supports reading, clicking, typing, tab management, and full-page screenshots.
  • 04Captures requests, responses, console output, and performance data for website debugging.
  • 05Exports loaded tables and lists as JSON or CSV.
  • 06Offers a server-side agent mode with the browser kept on your computer.

ユースケース

  • Having an agent read or summarize pages behind your company login.
  • Filling in web forms or completing multi-step workflows in your own accounts.
  • Reproducing a failing request on a website and collecting the evidence for debugging.
  • Exporting data tables from a web app into CSV for analysis.

トピック

#ai-agent#browser-automation#browser-extension#cli#chrome#edge#web-debugging#rust#agent#browser-use

よくある質問

BrowserSkill とは？
CLI and browser extension that let AI agents work in your logged-in Chrome or Edge while you keep using your browser. BrowserSkill connects a shell-capable AI agent to Chrome or Microsoft Edge, using the accounts you are already signed into. Agent tasks run in a separate, visible Agent Window, and an existing tab can be explicitly borrowed and returned when the task ends. Agents can read pages, click and type, manage tabs, capture screenshots, and upload or download files in local mode.
BrowserSkill は無料で使えますか？ライセンスは？
BrowserSkill は MIT ライセンスのオープンソースです。商用製品で使う前にライセンス条件を確認してください。
BrowserSkill は今もメンテナンスされていますか？
活発にメンテナンス中。BrowserSkill の最終コミットは 2026年10月15日 です。
BrowserSkill は何に使えますか？
Having an agent read or summarize pages behind your company login. Filling in web forms or completing multi-step workflows in your own accounts. Reproducing a failing request on a website and collecting the evidence for debugging. Exporting data tables from a web app into CSV for analysis.
BrowserSkill はどこで入手できますか？
ソースコードとインストール手順は GitHub にあります：https://github.com/Tencent/BrowserSkill

概要はプロジェクトの README をもとに AI が生成しています。GitHub の数値は毎日更新されます。

関連プロジェクト