← 返回 Skills 市场
Playwright MCP
作者
Spiceman161
· GitHub ↗
· v1.0.0
43188
总下载
155
收藏
552
当前安装
1
版本数
在 OpenClaw 中安装
/install playwright-mcp
功能描述
Browser automation via Playwright MCP server. Navigate websites, click elements, fill forms, extract data, take screenshots, and perform full browser automation workflows.
安全使用建议
Install only if you want an agent to control a browser through Playwright MCP. Prefer host allow-lists, avoid unsupervised use on logged-in or high-value accounts, and review actions that submit forms, upload files, make purchases, change settings, or run page JavaScript.
功能分析
Type: OpenClaw Skill
Name: playwright-mcp
Version: 1.0.0
The OpenClaw AgentSkills skill bundle for 'playwright-mcp' is classified as benign. The `SKILL.md` file provides clear documentation for browser automation using Playwright MCP, including installation instructions and usage examples, without any evidence of prompt injection attempts or instructions for malicious actions. It even highlights built-in security features like host validation and restricted file system access. The `examples.py` script is a conceptual demonstration of the skill's usage and does not perform any actual harmful operations, using `example.com` as a placeholder and explicitly stating its conceptual nature. There is no indication of data exfiltration, persistence, or other malicious intent across any of the files.
能力评估
Purpose & Capability
Navigation, clicking, typing, screenshots, file upload, and page JavaScript execution are high-impact browser capabilities, but they directly match the stated Playwright MCP automation purpose and are disclosed.
Instruction Scope
The instructions are reference-style usage examples for starting and using the MCP server; no prompt override, hidden agent directives, or unrelated workflows were found.
Install Mechanism
Installation uses the disclosed external npm package @playwright/mcp through npm or npx. That is coherent for this skill, though the package version is not pinned.
Credentials
Browser automation can interact with logged-in sites and expose page contents through screenshots, traces, videos, or extraction scripts, so users should scope targets carefully.
Persistence & Privilege
Artifacts show a user-started MCP server and a conceptual example script that prints sample calls; no auto-start persistence, privilege escalation, credential harvesting, or local session-store access was found.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install playwright-mcp - 安装完成后，直接呼叫该 Skill 的名称或使用
/playwright-mcp触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release
元数据
常见问题
Playwright MCP 是什么？
Browser automation via Playwright MCP server. Navigate websites, click elements, fill forms, extract data, take screenshots, and perform full browser automation workflows. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 43188 次。
如何安装 Playwright MCP？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install playwright-mcp」即可一键安装，无需额外配置。
Playwright MCP 是免费的吗？
是的，Playwright MCP 完全免费（开源免费），可自由下载、安装和使用。
Playwright MCP 支持哪些平台？
Playwright MCP 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（linux, darwin, win32）。
谁开发了 Playwright MCP？
由 Spiceman161（@spiceman161）开发并维护，当前版本 v1.0.0。
推荐 Skills