菜单
Enable browser automation capabilities in Claude Code using the Chrome extension. This allows agents to interact with web pages, fill forms, click buttons, and perform automated browser tasks.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/stevehuang0115/crewly --skill claude-code-chrome-browser命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Assign a task to a worker within the Team Leader's subordinate scope. Validates that the target worker's parentMemberId matches the TL's memberId before delegation. Includes auto-monitoring setup.
Create a new task via the task-management API for autonomous work decomposition.
Mark a task as blocked with a reason and optional questions for the orchestrator.
基于 SOC 职业分类
正在显示 SKILL.md
| name | Claude Code Chrome Browser |
| description | Enable browser automation capabilities in Claude Code using the Chrome extension. This allows agents to interact with web pages, fill forms, click buttons, and perform automated browser tasks. |
| version | 1.0.0 |
| category | automation |
| skillType | mcp |
| triggers | ["browser","chrome","web automation","browse","click","fill form","navigate"] |
| tags | ["browser","automation","chrome","mcp","web"] |
| execution | {"type":"mcp-tool","mcpTool":{"toolName":"mcp__claude-in-chrome__*","defaultParams":null}} |
| runtime | {"runtime":"claude-code","flags":["--chrome"]} |
| notices | [{"type":"requirement","message":"This skill requires the Claude Code Chrome extension to be installed in your browser."},{"type":"info","message":"This skill is only available when using Claude Code as the AI runtime."}] |
This skill enables browser automation capabilities through the Claude Code Chrome extension.
When this skill is enabled, agents can:
The browser automation tools are available through MCP (Model Context Protocol) when Claude Code is started with the --chrome flag.
mcp__claude-in-chrome__navigate - Navigate to URLsmcp__claude-in-chrome__read_page - Read page accessibility treemcp__claude-in-chrome__find - Find elements using natural languagemcp__claude-in-chrome__form_input - Fill form fieldsmcp__claude-in-chrome__computer - Mouse and keyboard interactionsmcp__claude-in-chrome__javascript_tool - Execute JavaScriptmcp__claude-in-chrome__get_page_text - Extract page text content--chrome flag (Crewly handles this automatically when this skill is enabled)find to locate elements by description before clicking