← Back to Skills Marketplace
229
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install page-js-operation
Description
通过 CDP 和 Page JS Extension 操作网页元素，节省 95% Token（仅用于 UI 自动化测试）
Usage Guidance
This skill is internally consistent for UI automation, but it depends on a third-party Chrome extension you must trust. Before installing: (1) review the extension repo and manifest.json for any network or background permissions (host_permissions, webRequest, remote code loads); (2) load the extension into an isolated Chrome profile used only for testing; (3) restrict which domains the extension can access; (4) ensure your AI agent is configured so it will not send element keys to untrusted external LLM endpoints (or disable network access during tests); (5) avoid using the skill on any page with sensitive data. If you cannot or will not review the extension source, treat the extension as untrusted and do not install it in a profile with real accounts or sensitive data.
Capability Analysis
Type: OpenClaw Skill
Name: page-js-operation
Version: 1.0.2
The skill facilitates browser automation by interfacing with a third-party Chrome extension via the Chrome DevTools Protocol (CDP). While it includes extensive safety warnings and explicitly prohibits use on sensitive sites, it introduces high-risk capabilities including the execution of arbitrary JavaScript and the requirement to install external code from a GitHub repository (TangJing/openclaw_access_web_page_js). Furthermore, the skill's core logic involves extracting DOM metadata (including innerText) and sending it to the LLM, which presents a significant risk of data exposure if the agent is used on pages containing sensitive information.
Capability Assessment
Purpose & Capability
The skill's name/description (token-efficient web operations for UI automation) matches what it asks for: a Page JS Chrome extension and CDP access. There are no unrelated env vars, binaries, or hidden requirements.
Instruction Scope
SKILL.md confines actions to using ElementCollector via CDP (exporting element keys, searching, clicking, filling non-sensitive inputs). It explicitly warns against sensitive pages and discloses that element keys may include innerText. Note: the extension can be manipulated via native DOM APIs to write to non-indexed fields (e.g., password/hidden/file), and the agent may send element keys to external LLMs — so follow the prohibition on sensitive pages and ensure agent behaviour is constrained.
Install Mechanism
Instruction-only skill (no automated install). The extension must be manually downloaded from an external GitHub repo and loaded unpacked. That reduces platform-level risk, but the extension repo is unverified: you must review manifest.json and source for network calls/remote code before loading. No automated downloads or extraction are performed by the skill itself.
Credentials
Skill declares no credentials, env vars, or config paths — appropriate for a local browser automation helper. The only data flow risk is that element keys (which may contain page text) could be sent by the AI agent to LLM services; the skill flags this, but protecting that requires agent configuration.
Persistence & Privilege
No elevated persistence is requested (always:false). The skill does not modify other skills or system configs. Autonomous invocation is enabled by default on the platform, which is normal; consider restricting autonomous runs if you want tighter control.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install page-js-operation - After installation, invoke the skill by name or use
/page-js-operation - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
修改隐私协议，并修改隐私规范，增加英文版本
v1.0.1
- 增加了“隐私说明”板块，强调所有插件运行及数据处理均在本地浏览器进行，无数据传出或隐私泄露风险
- 说明插件的本地运行、无网络请求、内存数据特性，并阐明设计初衷是为节省 LLM Token
- 明确指出插件代码完全开源，用户可自由下载源码安装
v1.0.0
must install chrome extensions;
Metadata
Frequently Asked Questions
What is Token Efficient Web Operations?
通过 CDP 和 Page JS Extension 操作网页元素，节省 95% Token（仅用于 UI 自动化测试）. It is an AI Agent Skill for Claude Code / OpenClaw, with 229 downloads so far.
How do I install Token Efficient Web Operations?
Run "/install page-js-operation" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Token Efficient Web Operations free?
Yes, Token Efficient Web Operations is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Token Efficient Web Operations support?
Token Efficient Web Operations is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Token Efficient Web Operations?
It is built and maintained by Jing Tang (@tangjing); the current version is v1.0.2.
More Skills