BrowserSkill让Agent操作已登录浏览器

Let AI agents use your real, logged-in browser without interrupting your work. CLI + extension for browser automation across any shell-ca…

BrowserSkill Let AI agents work in your logged-in browser while you keep working. English · 中文 Quick start · Website debugging · DSH plugin · Documentation · Changelog BrowserSkill connects your AI agent to Chrome or Microsoft Edge, using the accounts you are already signed into. Ask it to read pages, fill forms, work through a website, capture a long screenshot, or investigate a failing request. Tasks run in a separate, visible Agent Window; an existing tab can be explicitly borrowed and returned when the task ends. Use the bsk CLI with Cursor, Claude Code, Codex, OpenClaw, CodeBuddy, WorkBuddy, Pi, Hermes Agent, or another shell-capable agent. DeepSeek Harness has a dedicated plugin with native browser tools and task previews. You choose the agent and model; BrowserSkill provides the browser connection. What you can do Capability What it gives you Work with your existing accounts Use the browser's current login state to read documents, search internal sites, fill forms, and complete web workflows. Keep browser tasks visible Give the agent its own window, borrow an existing tab when needed, and take over for login, verification, or other human-only steps. Read, interact, and capture Inspect page text and controls, click and type, manage tabs, capture viewport or full-page screenshots, and upload or download files in local mode. Debug websites with evidence Connect actions to requests, response bodies, console messages, and page changes. Inspect performance, slow APIs, and suspected duplicate requests; use explicit HTTP rules or request replay to test a hypothesis. Choose the right browser Name browser instances, bind a task to a specific profile, or pair an agent running on a server with a browser on your computer. Review what happened Reopen browser-local debugging history, export evidence as JSON, or enable a separate operation audit to review task activity. Extract structured data Export loaded tables and lists as JSON/CSV with column names, page/frame sources, merged-cell metadata and explicit coverage. See structured extraction. Record a task as video Ask the agent to record one authorized tab as a silent MP4, then preview, save or delete the video from the extension. See task video recording. Wait for page conditions Wait for a requested element condition with an explicit timeout before continuing a workflow. See element waits. Watch a browser task in action: github.-.mp4 Quick Start For local automation, you need an AI agent + the bsk CLI + the browser extension. The CLI includes the background daemon. The skill teaches your agent how to use it. Component Supported environments CLI / daemon macOS: Apple Silicon and Intel · Linux: x64 and ARM64 · Windows: x64 Extension Chrome and Microsoft Edge, based on Chromium 125 or later. Other Chromium browsers may work; compatibility is not guaranteed. Agent integration A shell-capable agent with the BrowserSkill skill, or DeepSeek Harness with the DSH plugin. Let your agent set it up Send this to your agent: Set up browser-skill on this machine by following https://raw.githubusercontent.com/Tencent/BrowserSkill/main/AGENT_INSTALL.md The guide covers CLI installation, the correct skill or DSH plugin, connection checks, and a first browser task. You will still need to install the extension in the browser you want to use: Install for Chrome · Install for Edge Manual setup 1. Install the CLI macOS / Linux: curl -fsSL https://raw.githubusercontent.com/Tencent/BrowserSkill/main/install.sh | sh export PATH="${BSK_INSTALL_DIR:-$HOME/.local/bin}:$PATH" Windows PowerShell: irm https://raw.githubusercontent.com/Tencent/BrowserSkill/main/install.ps1 | iex By default, the binary is installed under ~/.local/bin. Check it in the terminal or agent environment that will use it: bsk --version If an already-running agent cannot find bsk, restart it to pick up the new PATH, or configure the installed binary's absolute path. 2. Connect the extension Install it from the Chrome Web Store or Edge Add-ons. Open its popup, enable the local connection, and check its status after starting the CLI. 3. Install the skill for your agent bsk install-skill Select your harness with Space, then press Enter. For non-interactive setup, choose it explicitly: bsk install-skill --harness cursor --json Use bsk install-skill --list to see supported targets and paths. Existing installations are skipped unless you explicitly replace them with --force. For another harness, copy the entire crates/bsk-cli/skill/ directory into its skills directory as browser-skill/, including references/. DSH users should install the plugin instead; it includes the skill. 4. Check the connection bsk doctor Resolve any failed checks and confirm that the extension shows Connected. Start a new agent session and check that it can discover browser-skill; a successful doctor check alone does not confirm skill discovery. Try your first task Once connected, ask your agent: Use browser-skill to open https://example.com, summarize the page, and end the browser session when finished. The agent opens its own window in the background, reads the page, returns the summary, and stops its task. External links or manually created tabs that land in the task window are moved to your most recently used normal window; a user window is created if none remain. Routing happens after tab creation, so a tab may briefly appear in the task window. Page-opened tabs and human assistance or recording are preserved. Harnesses that support skill slash commands can also use /browser-skill. Chrome can still activate a window when a page opens a native popup. For observed task popups, the extension restores the previous Chrome window or minimized state if you have not switched windows. Chrome's public extension APIs cannot guarantee zero transient activation or restore focus to another application. Try the CLI directly Start a session and keep the returned session_id: bsk session start --json Replace <id> with that value in each command be