← Back to Skills Marketplace
c595390153-a11y

Cn Web Search Custom

by c595390153-a11y · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
73
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install cn-web-search-custom
Description
中文网页搜索 - 聚合 17 个免费搜索引擎，无需 API Key，纯网页抓取，支持公众号/财经/技术/学术/知识搜索。
README (SKILL.md)

中文网页搜索 (CN Web Search)

⚡ 安装:

clawhub install cn-web-search

多引擎聚合搜索，全部免费，无需 API Key，纯网页抓取。17 个引擎覆盖中英文、公众号、技术、财经、知识百科。

引擎总览（17 个）

类别 引擎 数量
公众号 搜狗微信、必应索引 2
中文综合 360、搜狗、必应中文、百度、头条搜索 5
英文综合 DDG Lite、Qwant、Startpage、必应英文、Yahoo、Brave Search、Mojeek 7
技术社区 Stack Overflow、GitHub Trending 2
财经/投资 东方财富、集思录、财新 3
知识百科 Wikipedia 中文、Wikipedia 英文 2

全部通过 web_fetch 抓取网页，零 API 依赖。

1. 公众号搜索

1.1 搜狗微信

https://weixin.sogou.com/weixin?type=2&query=QUERY&page=1

1.2 必应公众号索引

https://cn.bing.com/search?q=site:mp.weixin.qq.com+QUERY

2. 中文综合搜索

2.1 360 搜索

https://m.so.com/s?q=QUERY

2.2 搜狗网页

https://www.sogou.com/web?query=QUERY

2.3 必应中文

https://cn.bing.com/search?q=QUERY

2.4 百度

https://www.baidu.com/s?wd=QUERY

中文搜索覆盖最全，结果丰富。

2.5 头条搜索

https://so.toutiao.com/search?keyword=QUERY

字节跳动旗下，中文资讯和短视频内容强。

3. 英文综合搜索

3.1 DuckDuckGo Lite

https://lite.duckduckgo.com/lite/?q=QUERY

3.2 Qwant

https://www.qwant.com/?q=QUERY&t=web

3.3 Startpage

https://www.startpage.com/do/search?q=QUERY&cluster=web

3.4 必应英文

https://www.bing.com/search?q=QUERY

3.5 Yahoo

https://search.yahoo.com/search?p=QUERY

老牌英文搜索引擎，结果稳定。

3.6 Brave Search

https://search.brave.com/search?q=QUERY

独立索引（非 Bing/Google 代理），隐私友好，结果质量高。

3.7 Mojeek

https://www.mojeek.com/search?q=QUERY

独立爬虫索引，不依赖任何大厂，适合多样化结果。

4. 技术社区

4.1 Stack Overflow

https://stackoverflow.com/search?q=QUERY

4.2 GitHub Trending

https://github.com/trending?since=weekly

5. 财经/投资

5.1 东方财富

https://search.eastmoney.com/search?keyword=QUERY

5.2 集思录

https://www.jisilu.cn/explore/?keyword=QUERY

投资社区，可转债、基金、LOF 等投资品种讨论。

5.3 财新

https://search.caixin.com/search/?keyword=QUERY

6. 知识百科

6.1 Wikipedia 中文

https://zh.wikipedia.org/w/index.php?search=QUERY&title=Special:Search

中文维基百科，知识查询首选。

6.2 Wikipedia 英文

https://en.wikipedia.org/w/index.php?search=QUERY&title=Special:Search

英文维基百科，信息量最大的免费百科全书。

使用示例

中文搜索：

  • 百度：web_fetch(url="https://www.baidu.com/s?wd=英伟达财报", extractMode="text", maxChars=12000)
  • 360：web_fetch(url="https://m.so.com/s?q=英伟达财报", extractMode="text", maxChars=12000)

英文搜索：

  • Brave：web_fetch(url="https://search.brave.com/search?q=AI+news", extractMode="text", maxChars=8000)
  • DDG：web_fetch(url="https://lite.duckduckgo.com/lite/?q=AI+news", extractMode="text", maxChars=8000)

公众号：

  • 搜狗微信：web_fetch(url="https://weixin.sogou.com/weixin?type=2&query=英伟达&page=1", extractMode="text", maxChars=10000)

知识查询：

  • Wikipedia：web_fetch(url="https://zh.wikipedia.org/w/index.php?search=量子计算&title=Special:Search", extractMode="text", maxChars=8000)

投资：

  • 集思录：web_fetch(url="https://www.jisilu.cn/explore/?keyword=可转债", extractMode="text", maxChars=8000)

引擎选择建议

场景 推荐引擎
中文通用搜索 百度 → 360 → 搜狗
英文通用搜索 Brave → DDG → Bing
公众号文章 搜狗微信 → 必应索引
技术问题 Stack Overflow → GitHub
最新资讯 头条搜索 → 百度
A股/投资 东方财富 → 集思录
财经深度 财新
知识/定义 Wikipedia 中文 → Wikipedia 英文
隐私优先 Brave → Mojeek → DDG

实战对比：有 cn-web-search vs 没有

以投研场景为例，问同一个问题："英伟达2026年Q1财报业绩如何？"

❌ 没有 cn-web-search（纯模型）

回答："我的训练数据有截止日期，无法提供最新财报数据。
      建议您查看英伟达官网 investor.nvidia.com。"

结果：什么有用信息都拿不到。

✅ 装了 cn-web-search（百度 + 360 + 搜狗，3 个免费引擎）

web_fetch(url="https://www.baidu.com/s?wd=英伟达2027财年Q1业绩指引", extractMode="text", maxChars=3000)
web_fetch(url="https://m.so.com/s?q=英伟达2026财报Q1业绩", extractMode="text", maxChars=3000)
web_fetch(url="https://www.sogou.com/web?query=英伟达财报2026Q1业绩预测", extractMode="text", maxChars=3000)

拿到的实时数据（多源交叉验证）：

数据点 百度 360 搜狗
FY2026 Q1 营收 441 亿美元 (+69%) ✅ ✅ ✅
FY2026 Q4 营收 681 亿美元 (+73%) ✅ ✅ ✅
FY2026 全年营收 2159 亿美元 (+65%) ✅ ✅ ✅
FY2027 Q1 指引 780 亿（超预期 7%） ✅ ✅ ✅
H20 禁令影响 80 亿美元 ✅ ✅ ✅
毛利率 75% ✅ ✅ —
黄仁勋："代理式 AI 拐点已到来" ✅ — ✅

对比总结

维度 无 Skill cn-web-search
实时数据 ❌ ✅
数据准确性 N/A ✅ 多源交叉验证
成本 — 免费
投研可用性 零 营收/利润/指引/管理层表态
信息来源 无 百度+360+搜狗+雪球+东方财富

结论：没有 cn-web-search 的 agent 在投研场景下基本是废的。装上后等于给 agent 开了一扇窗，能看到实时世界——而且完全免费。

更新日志

v2.2.0

  • 📊 新增「实战对比」章节：有 cn-web-search vs 没有（以英伟达财报投研为例）
  • 📈 展示多源交叉验证效果

v2.1.0

  • 🗑️ 移除所有 API 端点引擎（Hacker News、Reddit、ArXiv、DDG API、Wolfram Alpha）
  • ✅ 全部 17 个引擎均为纯网页抓取，零 API 依赖
  • 📋 更新引擎总览表和选择建议

v2.0.0

  • 🆕 新增 9 个引擎：百度、Yahoo、Brave Search、Mojeek、头条搜索、集思录、Wikipedia 中英文、DDG Instant Answer API
  • 📊 引擎总数：13 → 22

v1.0.0

  • ✅ 全新发布！聚合 13+ 免费中文搜索引擎
  • ✅ 无需 API Key，真正免费
  • ✅ 支持公众号、知乎、财经(A股)、技术搜索
Usage Guidance
This skill appears to be what it says: a set of instructions to fetch public search result pages. Before installing: 1) Verify the author/repository (SKILL.md references github.com/joansongjr) — metadata mismatches suggest the package may have been copied or repackaged. 2) Be aware the skill causes the agent to make network requests to many third-party sites (public search engines); consider privacy, scraping policies, rate limits, and legal/terms-of-service for those sites. 3) Test in a sandbox or with limited invocation first to observe network behavior. 4) If you require provenance, ask the publisher to fix the metadata (slug, version, ownerId) or provide the canonical GitHub repo/release used to produce this package. 5) No credentials are required, but if you plan to combine this with other skills that have secrets, avoid letting the aggregator post sensitive data into fetched queries.
Capability Analysis
Type: OpenClaw Skill Name: cn-web-search-custom Version: 1.0.0 The skill bundle is a collection of search engine URL templates and instructions for an AI agent to perform web searches using its built-in 'web_fetch' tool. It covers 17 engines including Baidu, Bing, and Brave Search without requiring API keys. There is no executable code, evidence of data exfiltration, or malicious prompt injection; the instructions in SKILL.md are strictly aligned with the stated purpose of multi-engine searching and data verification.
Capability Tags
requires-sensitive-credentials
Capability Assessment
ℹ Purpose & Capability
The declared purpose (aggregate 17 free search engines via webpage scraping) matches the runtime instructions (explicit URLs and web_fetch examples). No unrelated credentials, binaries, or install steps are requested. However, there are metadata inconsistencies: registry slug is 'cn-web-search-custom' while package.json/_meta.json use 'cn-web-search', versions differ (registry 1.0.0 vs SKILL.md 2.3.0 vs package.json 0.9.0), and ownerId in _meta.json differs from top-level owner ID; these suggest sloppy packaging or copy-paste rather than a functional mismatch, but they lower confidence and merit verification.
✓ Instruction Scope
SKILL.md only instructs the agent to call web_fetch on public search URLs and gives engine-selection guidance. It does not instruct reading local files, environment variables, or sending data to third-party endpoints other than the listed search engines. The primary runtime action is network requests to public websites, which is consistent with the stated purpose.
✓ Install Mechanism
No install spec or code files are required; this is an instruction-only skill. That minimizes disk-write and code-execution risk. The README/package metadata reference a GitHub repo and a 'clawhub install' example, but no installer is bundled with the skill package here.
✓ Credentials
The skill requests no environment variables, no credentials, and no config paths. Examples use web_fetch only. There is no disproportionate access requested.
✓ Persistence & Privilege
Flags show always:false and default agent invocation is allowed (normal). The skill does not request persistent/privileged settings or change other skills' configurations. Note: as with any web-fetching skill, autonomous invocation allows the agent to make outbound HTTP requests at runtime.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install cn-web-search-custom
  3. After installation, invoke the skill by name or use /cn-web-search-custom
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
cn-web-search-custom 1.0.0 - Initial release: aggregate search across 13+ free Chinese search engines. - No API key required; all engines are free to use. - Supports multi-category search: WeChat Official Accounts, Zhihu, finance (A-shares), technical search, and more. - Pure web scraping approach for fetching results.
Metadata
Slug cn-web-search-custom
Version 1.0.0
License MIT-0
All-time Installs 1
Active Installs 1
Total Versions 1
Frequently Asked Questions

What is Cn Web Search Custom?

中文网页搜索 - 聚合 17 个免费搜索引擎，无需 API Key，纯网页抓取，支持公众号/财经/技术/学术/知识搜索。 It is an AI Agent Skill for Claude Code / OpenClaw, with 73 downloads so far.

How do I install Cn Web Search Custom?

Run "/install cn-web-search-custom" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Cn Web Search Custom free?

Yes, Cn Web Search Custom is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Cn Web Search Custom support?

Cn Web Search Custom is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Cn Web Search Custom?

It is built and maintained by c595390153-a11y (@c595390153-a11y); the current version is v1.0.0.

More Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463815 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259802 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200680 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191345 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190333 · by oswalpalash

💬 Comments