← 返回 Skills 市场
hkall

Easy Search

作者 hkall · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ 安全检测通过
524
总下载
0
收藏
12
当前安装
2
版本数
在 OpenClaw 中安装
/install easy-search
功能描述
Simple web search using multiple search engines with no API key required. Supports Google, Bing, DuckDuckGo, Baidu, Sogou, 360, Brave, Yandex. Features time...
使用说明 (SKILL.md)

Easy Search Skill

A powerful web search skill that doesn't require API keys. It uses direct HTTP requests and the duckduckgo_search library for reliable results.

Features

  • No API key required - Uses public search interfaces
  • 9 search engines - Google, Bing, DuckDuckGo, Baidu, Sogou, 360, Brave, Yandex, Startpage
  • Auto engine selection - Automatically picks the best available engine
  • Snippet extraction - Get result previews, not just titles
  • Time filtering - Filter by day/week/month/year
  • Interactive mode - Continuous searching with live engine switching
  • Network diagnostics - Built-in connectivity checking and proxy hints
  • Proxy support - Respects ALL_PROXY, HTTPS_PROXY environment variables
  • Result caching - Faster repeated searches
  • Smart failover - Auto-switches to backup engines on failure

Requirements

  • Python 3.6+
  • Optional: pip install duckduckgo-search requests for better results

Commands

Basic Search

# Simple search (positional argument)
python3 {baseDir}/scripts/search.py Python tutorial

# With flag
python3 {baseDir}/scripts/search.py --query "your search terms"

# Specify engine
python3 {baseDir}/scripts/search.py "AI news" --engine bing

# More results
python3 {baseDir}/scripts/search.py "React hooks" --results 10

# JSON output (default)
python3 {baseDir}/scripts/search.py "Vue.js" --format json

# Markdown output
python3 {baseDir}/scripts/search.py "machine learning" --format md

# Simple output (quick reading)
python3 {baseDir}/scripts/search.py "Docker" --format simple

# CSV output
python3 {baseDir}/scripts/search.py "Kubernetes" --format csv

Time Filtering

# News from last 24 hours
python3 {baseDir}/scripts/search.py "AI news" --time day

# Results from last week
python3 {baseDir}/scripts/search.py "React 19" --time week

# Results from last month
python3 {baseDir}/scripts/search.py "Python 3.13" --time month

# Results from last year
python3 {baseDir}/scripts/search.py "best laptops 2024" --time year

Interactive Mode

# Start interactive mode
python3 {baseDir}/scripts/search.py --interactive

# In interactive mode:
# [auto]> Python tutorial          # Search with auto engine
# [auto]> :engine duckduckgo       # Change engine
# [duckduckgo]> React hooks        # Search with new engine
# [duckduckgo]> :format simple     # Change output format
# [duckduckgo]> :check             # Check engine health
# [duckduckgo]> :quit              # Exit

Network Diagnostics

# Run network diagnostics
python3 {baseDir}/scripts/search.py --diagnose

# Check which engines are available
python3 {baseDir}/scripts/search.py --check-engines

Engine Management

# Clear cache
python3 {baseDir}/scripts/search.py --clear-cache

# Disable auto-fallback (stay on specified engine)
python3 {baseDir}/scripts/search.py "query" --engine google --no-fallback

Search Engines

Engine Best For Notes
auto General use Auto-selects best available engine
startpage Privacy, global Privacy-focused, reliable HTML parsing
duckduckgo Privacy, global Preferred, uses library if installed
bing Global search Good reliability, decent snippets
google Best results May need proxy in some regions
baidu Chinese content Strong anti-crawl measures
sogou Chinese content Good alternative to Baidu
so360 Chinese content Another Chinese option
brave Privacy-focused Growing index
yandex Russian/Global Works in some regions

Output Formats

JSON

{
  "query": "search terms",
  "engine": "bing",
  "method": "HTML scraping",
  "results": [
    {
      "title": "Result Title",
      "url": "https://example.com",
      "snippet": "Brief description..."
    }
  ],
  "total": 5
}

Markdown

## Search Results: your terms

Engine: **BING**
Method: _HTML scraping_

1. [Result Title](https://example.com)
   > Brief description...

2. [Another Title](https://another.com)
   > More info...

Simple

1. Result Title
   https://example.com
   Brief description...

2. Another Title
   https://another.com
   More info...

Examples

# Quick search for documentation
python3 {baseDir}/scripts/search.py "Python requests library docs"

# Search Chinese content with Sogou
python3 {baseDir}/scripts/search.py "人工智能最新进展" --engine sogou

# Get recent news in markdown format
python3 {baseDir}/scripts/search.py "OpenAI news" --time day --format md

# Use interactive mode for multiple searches
python3 {baseDir}/scripts/search.py -i

# Search with time filter
python3 {baseDir}/scripts/search.py "React 19 features" --time month --results 10

Proxy Configuration

Set environment variables for proxy support:

# Set proxy
export ALL_PROXY="http://127.0.0.1:7890"
export HTTPS_PROXY="http://127.0.0.1:7890"

# Then run search
python3 {baseDir}/scripts/search.py "your query"

Troubleshooting

Problem Solution
Empty results Try a different engine with --engine
Connection timeout Check proxy settings or use --timeout 60
Rate limiting Wait between searches or switch engines
DuckDuckGo slow Install library: pip install duckduckgo-search
No snippets Some engines have limited snippet extraction
Google blocked Use proxy or switch to Bing/DuckDuckGo
All engines fail Run --diagnose to check network

Tips for Best Results

  1. Use auto engine - Let the tool pick the best available engine
  2. Install dependencies - pip install duckduckgo-search requests for better reliability
  3. Use time filters - For news and recent topics, use --time day or --time week
  4. Try Chinese engines - Use sogou or so360 for Chinese content
  5. Interactive mode - Great for multiple related searches
  6. Cache is your friend - Repeated searches are instant (1 hour cache)
  7. Run diagnostics - Use --diagnose to troubleshoot network issues

Version History

  • v1.1.0 - Added Startpage engine with proper parsing, fixed HTML entity decoding in URLs
  • v1.0.1 - Added network diagnostics, better proxy hints, improved error messages
  • v1.0.0 - Added 4 new engines (Sogou, 360, Brave, Yandex), time filtering, interactive mode, snippet extraction, engine health caching
  • v4.0.0 - Integrated duckduckgo_search library, auto engine selection
  • v1.0.0 - Initial release with basic search
安全使用建议
This skill appears to do what it claims: it scrapes public search engine HTML, makes network requests, and stores a small cache in ~/.easysearch_cache. Before installing, consider: (1) don't use it to search for sensitive secrets because queries go to external search engines; (2) scraping may be brittle or blocked by some engines and could raise TOS concerns; (3) the tool respects proxy env vars — if you use a corporate or local proxy, your queries will go through it; (4) if you want extra assurance, review the full scripts locally or run them in a sandboxed/non-privileged account. Overall the skill is internally consistent and does not request disproportionate credentials or system access.
功能分析
Type: OpenClaw Skill Name: easy-search Version: 1.0.1 The 'easy-search' skill is a functional web search utility designed to scrape results from multiple search engines (Google, Bing, DuckDuckGo, etc.) without requiring API keys. The implementation in `scripts/search.py` uses standard Python libraries for network requests and regex-based HTML parsing, featuring legitimate capabilities such as result caching in the user's home directory, multi-threaded engine health checks, and proxy support. No evidence of malicious intent, data exfiltration, or prompt injection was found.
能力评估
Purpose & Capability
The name/description (multi-engine web search without API keys) matches the included script and SKILL.md. The code implements HTML scraping, optional duckduckgo_search usage, engine selection, snippet extraction, time filters, proxy support, and local caching — all of which are expected for this functionality.
Instruction Scope
Runtime instructions tell the agent to run scripts/search.py and to set proxy env vars if needed. The code performs network requests to many search engines, can run interactively and perform diagnostics, and writes cache files under the user's home (~/.easysearch_cache). The SKILL.md does not declare the cache path explicitly; aside from local caching and contacting search engines, there are no instructions to read unrelated system files or exfiltrate data.
Install Mechanism
There is no install spec and no external downloads; this is effectively an instruction+script bundle. That is low-risk from an install-supply perspective (nothing is being fetched or extracted at install time).
Credentials
The skill declares no required credentials or special environment variables. It does honor common proxy variables (ALL_PROXY, HTTPS_PROXY, etc.), which is appropriate for a network tool. No unrelated secrets or external credentials are requested.
Persistence & Privilege
The skill is not forced-always, does not ask to modify other skills, and only writes to its own cache directory in the user's home. It does not request elevated privileges or system-wide configuration changes.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install easy-search
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /easy-search 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.1
Version 1.0.1 - Added network diagnostics tools and enhanced proxy configuration guidance. - Improved error messages for failed searches and connectivity issues. - Updated documentation with clearer troubleshooting steps and usage instructions. - No changes made to core search features or supported engines.
v1.0.0
- Initial release of Easy Search: simple web search without API keys. - Supports Google, Bing, DuckDuckGo, and Baidu search engines. - Parses and returns search results with titles, URLs, and snippets. - Allows JSON and Markdown output formats. - Supports proxy via ALL_PROXY and user agent rotation to avoid blocking.
元数据
Slug easy-search
版本 1.0.1
许可证 MIT-0
累计安装 12
当前安装数 12
历史版本数 2
常见问题

Easy Search 是什么?

Simple web search using multiple search engines with no API key required. Supports Google, Bing, DuckDuckGo, Baidu, Sogou, 360, Brave, Yandex. Features time... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 524 次。

如何安装 Easy Search?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install easy-search」即可一键安装,无需额外配置。

Easy Search 是免费的吗?

是的,Easy Search 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Easy Search 支持哪些平台?

Easy Search 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Easy Search?

由 hkall(@hkall)开发并维护,当前版本 v1.0.1。

💬 留言讨论