360 Secure SkillHub
360 安全技能中心

Oxylabs Web Search

该 Skill 提供 Oxylabs AI Studio 的两个子命令：search 按查询返回匹配页面的标题、URL，并可选附带页面正文；scrape 将单个 URL 的内容抓取为干净 Markdown。脚本为无依赖 Python，支持 geo 定位与可选的 JS 渲染，需要配置 Oxylabs AI Studio API 密钥。

oxylabs
oxylabs
热度 38.37
最近更新 2026-06-11 17:15:46
分类：搜索依赖：代码执行 / 第三方 API应用场景：研究 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add oxylabs-web-search
Metadata
name
oxylabs-web-search
description
Search the web and read web pages with Oxylabs AI Studio (alternative to Brave, DuckDuckGo or Tavily). Use this whenever the user wants to search the web, look something up, find current information or news, gather sources and links, or fetch the content of a specific page. It has two commands. `search` returns matching pages — titles, URLs, and optionally the page text. `scrape` pulls the content of a single URL as clean Markdown. It handles JavaScript-heavy sites and geo-targeting, has a free tier, and runs as dependency-free Python.
metadata
{"openclaw":{"requires":{"env":["OXYLABS_AI_STUDIO_API_KEY"]},"primaryEnv":"OXYLABS_AI_STUDIO_API_KEY","envVars":[{"name":"OXYLABS_AI_STUDIO_API_KEY","required":true,"description":"Oxylabs AI Studio API key."}]}}

Oxylabs Web Search & Data

One tool for getting things off the web — pick a subcommand:

  • search — find pages for a query (titles, URLs, and optional page text)
  • scrape — get the content of one URL

Setup

Get an Oxylabs AI Studio API key from https://aistudio.oxylabs.io/api-key. Pass API KEY to OpenClaw via one of following methods:

OpenClaw config (~/.openclaw/openclaw.json) (RECOMMENDED):

{
  "skills": {
    "entries": {
      "oxylabs-web-search": {
        "enabled": true,
        "apiKey": "YOUR_OXYLABS_AI_STUDIO_API_KEY"
      }
    }
  }
}

Export environment variable:

export OXYLABS_AI_STUDIO_API_KEY="YOUR_OXYLABS_AI_STUDIO_API_KEY"

Via .env file

For ~/.openclaw/.env, add the assignment:

OXYLABS_AI_STUDIO_API_KEY=YOUR_OXYLABS_AI_STUDIO_API_KEY

Usage

# search the web
python3 scripts/oxylabs.py search "<query>" -n 5
python3 scripts/oxylabs.py search "<query>" -n 3 --content   # + page text per result

# get one page as Markdown
python3 scripts/oxylabs.py scrape "<url>"

Common flags: --geo (two-letter ISO country) · --render-js (off by default; slower).