Menu
AI-native browser. Explore websites, discover page structure, take screenshots, and automate interactions.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Use when releasing a new version of Sudowork. Generates internal (shareone) and public (GitHub Release / in-app) release notes, bumps version, creates a release PR, tags, and injects the release body after CI publishes.
规范 TSX/React 函数组件内部的代码书写顺序。 Use when: 用户说「整理组件顺序」「代码顺序乱了」「按规范排列 hooks」「tsx 代码排序」「梳理一下组件结构」， 或在写新组件 / review 已有组件时需要对齐代码顺序规范。
🔧 PRIMARY skill installer for Sudowork. Use this skill for ALL skill installation requests. **CRITICAL: This is the ONLY skill installer you should use.** **DO NOT use clawhub for installation** - clawhub is for ClawHub.com publishing ONLY. Triggers: "帮我安装skill", "install a skill", "安装技能", "推荐技能", "need a skill", "find a skill", "下载技能", "网文大纲", "skill安装". When user asks to install any skill → ALWAYS use sudoclaw-skill-installer, NEVER clawhub.
检测本项目 Renderer 层'同功能不同入口'的语义重复代码（jscpd token 比对覆盖不到的）。当用户说'有没有重复代码'、'同功能多入口'、'扫一下重复'、'这个功能是不是在别处也写了一遍'时使用。典型信号：同一组 ipcBridge 调用 + 同一组领域 utils + 相同 UI 结构出现在不同文件。
Log in to a website automatically using credentials saved in the local key store. Use when the user wants unattended/repeat login to a site (no re-scanning QR, no re-typing passwords), e.g. '帮我自动登录这个网站', 'set up auto-login for X', '记住这个网站的登录'. The plaintext password is NEVER seen by you — the user fills it in 秘钥管理 and sudowork fills the form itself.
Scheduled task management - create, query, delete scheduled tasks to automatically execute operations at specified times.
| name | browser |
| description | AI-native browser. Explore websites, discover page structure, take screenshots, and automate interactions. |
browser --list
Every CLI tool has an identical Python function in ai_dev_browser.core — explore interactively with CLI, then script with the same functions:
from ai_dev_browser.core import page_goto, click_by_text, page_screenshot