国内浏览器搜索终极增强

国内12大引擎聚合搜索，自动去重排序，支持时效筛选，优先匹配垂直领域，提供精准且实时的本土化信息答案。

Security Scan

The skill is an instruction-only Python scraper that uses curl.exe to query search engines (primarily Bing); it does not request credentials or install extras, though its marketing claim of '12 engines' is overstated compared to the shipped code.

Purpose & Capability

Name/description promise (12 domestic engines, aggregation) is stronger than the actual implementation: the code practically only reliably scrapes Bing and includes parsers for 360 and Zhihu which may often fail. Rework notes in SKILL.md acknowledge the original overclaim. The requirement to call curl.exe on Windows is consistent with the code and the stated implementation method, but the script is Windows-centric and will not work without curl.exe on PATH.

Instruction Scope

SKILL.md instructs running run.py with query parameters; the script only issues HTTP requests to search engine URLs via curl.exe, parses HTML with regex, deduplicates by URL, and prints results to stdout. It does not read arbitrary local files, access environment secrets, or transmit data to third-party endpoints outside the targeted search engines. Behavior stays within the stated purpose (web scraping for search results).

Install Mechanism

No install spec is provided (instruction-only), so nothing is written to disk by an installer. The only external dependency is curl.exe on PATH; this is low install risk but does require the user to ensure a trusted curl.exe is available (Windows).

Credentials

The skill declares no environment variables, credentials, or config paths and the code does not access any secrets. There are no disproportionate credential requests.

Persistence & Privilege

The skill is not always-enabled and is user-invocable only; it does not modify other skills or system-wide configuration. It runs on demand and does not request elevated/system persistence.

Assessment

This skill is basically a small Python scraper that runs curl.exe to fetch search pages and parses them with regex. It does not ask for secrets or install additional packages, but note: (1) it only reliably returns Bing results—the advertised '12 engines' is overstated and the SKILL.md even notes this; (2) it requires curl.exe on Windows (ensure curl.exe is the expected, trusted binary); (3) because it issues HTTP requests from your machine, avoid running it on sensitive hosts or inside networks where making arbitrary outbound requests could expose internal services; (4) the full source is included—review it if you have special security constraints, or run it in a sandboxed environment if unsure.

browser-search-ultimate-cn - 国内搜索技能

实际功能描述

通过 Windows curl.exe 直接请求搜索引擎页面，正则提取搜索结果。输出核心答案摘要 + 详细结果列表。

引擎支持（2026-03-31 实测）

引擎	状态
Bing (cn.bing.com)	✅ 稳定可用
360 (so.com)	❌ 被反爬
知乎	❌ 被反爬

当前有效引擎：Bing（主力，稳定出结果）

适用场景

• 国内中文信息查询
• 新闻、热点事件
• 通用搜索

使用方式

直接调用 Python 脚本：

run.py <关键词> [freshness] [count] [vertical]

参数说明

参数	类型	必填	说明
query	string	是	搜索关键词
freshness	string	否	预留参数（暂未实现）
count	number	否	返回数量，默认10条
vertical	string	否	news/tech/ent/shop（引擎组合略有不同）

输出示例

== [Search] 中东战争 ==

[Bing] +10

== [Core Answer] ==
  摘要内容1
  摘要内容2

== [Results: 10 | Bing + 360 + Zhihu] ==
1. [Bing] 标题
   摘要
   URL

技术实现

• 纯 Python 标准库（sys, io, subprocess, re, urllib, collections）
• subprocess.run(['curl.exe', ...]) 发 HTTP 请求
• re.finditer 正则提取搜索结果
• OrderedDict 按 URL 去重
• 输出编码：sys.stdout = io.TextIOWrapper(..., encoding='utf-8') 解决 Windows GBK 问题

局限性

1. 依赖 curl.exe 在系统 PATH 中可用
2. 仅 Bing 稳定，其他引擎可能被反爬返回空结果
3. 不支持需要 JavaScript 渲染的页面（知乎、B站等）
4. 无广告过滤、无权威度排序
5. 摘要来自 HTML 原始文本，可能含噪声字符

修复记录

• 2026-03-31：重写 run.py，移除对不存在的 openclaw tool call CLI 命令的依赖，改为直接用 curl 请求 Bing
• 原版 skill 声称支持 12 引擎 + OpenClaw CLI 集成，与实际代码不符，已更正