AI Agent Skill HUB
multiAI Summary Pending

pwdebug

用于需要通过命令行操作真实浏览器实例进行前端调试（如导航、执行 JS、截图、元素拾取、控制台日志）且希望跨多次命令复用同一浏览器会话的场景。

231 stars
byaiskillstore
View on GitHubInstallation ↓

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/pwdebug/SKILL.md --create-dirs "https://raw.githubusercontent.com/aiskillstore/marketplace/main/skills/dcjanus/pwdebug/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/pwdebug/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How pwdebug Compares

Feature / AgentpwdebugStandard Approach
Platform SupportmultiLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

用于需要通过命令行操作真实浏览器实例进行前端调试（如导航、执行 JS、截图、元素拾取、控制台日志）且希望跨多次命令复用同一浏览器会话的场景。

Which AI agents support this skill?

This skill is compatible with multi.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Playwright 浏览器调试 CLI

## 概览

该技能提供一个基于 Playwright 的命令行工具，用于启动浏览器服务并执行导航、评估 JS、截图、元素拾取与日志监听等调试操作。当前仅支持 Chromium（通过 CDP 连接）。

## 快速开始

> 工作目录应为本文件所在目录，示例命令默认从该目录执行。

1. 启动浏览器服务（常驻进程）：

```bash
scripts/pwdebug.py start
```

2. 在新标签页打开页面：

```bash
scripts/pwdebug.py nav https://example.com --new
```

3. 执行 JS 表达式：

```bash
scripts/pwdebug.py evaluate "document.title"
```

4. 截图：

```bash
scripts/pwdebug.py screenshot --full
```

5. 交互式拾取元素：

```bash
scripts/pwdebug.py pick "点击登录按钮"
```

6. 监听控制台日志：

```bash
scripts/pwdebug.py watch-logs
```

7. 查看最近日志：

```bash
scripts/pwdebug.py logs 100
```

## 说明

- CLI 入口：`scripts/pwdebug.py`
- 日志路径：`~/.cache/pwdebug/console.log.jsonl`
- 状态路径：`~/.cache/pwdebug/server.json`

## 依赖与安装

- 脚本依赖通过 `uv --script` 管理。