Zhipu AI Search

使用智谱 Web Search API 进行网络搜索。当用户需要搜索最新信息、查找实时数据、查询新闻、研究某个话题或获取网络上的信息时使用。支持多种搜索引擎（智谱基础版/高阶版、搜狗、夸克），可配置搜索范围、时间筛选、结果数量等参数。

Security Scan

The skill's code and runtime instructions match its stated purpose (calling Zhipu/BigModel web search via an API key); only minor metadata mismatches exist in the registry entry.

Purpose & Capability

The skill is described as a web search wrapper for 智谱 (BigModel) and the shipped script (search_util.py), SKILL.md, and README implement exactly that (HTTP POST to open.bigmodel.cn web_search). However, the registry metadata claims no required environment variables while SKILL.md and the script require ZHIPUAI_API_KEY — a metadata inconsistency that should be corrected.

Instruction Scope

SKILL.md confines runtime behavior to obtaining a ZHIPUAI_API_KEY, constructing search requests, and printing structured results. It does not instruct reading unrelated files, scanning system state, or exfiltrating data to third-party endpoints; the only external endpoint is the documented BigModel API.

Install Mechanism

No install spec (instruction-only) is present, which minimizes risk. A small requirements.txt references httpx from PyPI — expected for making HTTP requests. There are no downloads from arbitrary URLs or archive extraction steps in the repo.

Credentials

The only secret the skill needs is ZHIPUAI_API_KEY, which is proportional to its stated purpose. Note again the registry metadata omission: the central registry entry lists no required env vars while SKILL.md and the script require the API key.

Persistence & Privilege

The skill does not request always:true and does not modify other skills or global agent settings. .claude/settings.json grants Bash/python invocation permissions (expected for a Python script). Autonomous invocation is allowed by default but is not unusually privileged here.

Assessment

This skill appears to do what it claims: call 智谱/BigModel's web_search API and present results. Before installing, be sure you are comfortable providing a ZHIPUAI_API_KEY (the script uses it in an Authorization header). Note the registry metadata did not list this required env var — verify the platform will prompt you to supply ZHIPUAI_API_KEY and not leave it undocumented. As with any third-party skill, only supply API keys you trust and consider using a limited-scope key if the provider supports it. Finally, install the dependency (httpx) in a controlled environment (virtualenv) before running the script.

智谱 Web Search Skill

使用智谱 AI 的 Web Search API 进行网络搜索，获取实时、结构化的搜索结果。

前置条件

1. 获取智谱 BigModel 开放平台的 API Key：https://bigmodel.cn/usercenter/apikeys
2. 配置 API Key（见下方说明）

API Key 配置

在使用前，请设置您的智谱 API Key：

export ZHIPUAI_API_KEY="your-api-key-here"

重要: 请使用环境变量方式配置 API Key，不要将密钥硬编码在脚本或配置文件中。

快速开始

使用提供的 search_util.py 脚本进行搜索：

# 基本搜索
python search_util.py "搜索内容"

# 查看帮助
python search_util.py --help

搜索引擎选项

引擎编码	说明
search_std	智谱基础版搜索引擎
search_pro	智谱高阶版搜索引擎（推荐）
search_pro_sogou	搜狗搜索
search_pro_quark	夸克搜索

搜索参数

参数	类型	必填	说明
search_query	String	是	搜索内容，建议不超过 70 字符
search_engine	String	是	搜索引擎编码
search_intent	Boolean	是	是否进行意图识别
count	Int	否	返回结果条数 (1-50, 默认 10)
search_recency_filter	String	否	时间筛选: oneDay/oneWeek/oneMonth/oneYear/noLimit
content_size	String	否	内容长度: medium/high
search_domain_filter	String	否	限定域名

API 概览

请求地址: https://open.bigmodel.cn/api/paas/v4/web_search

认证方式: 使用智谱 API Key，通过 Authorization header 传递

响应结构

{
  "id": "任务ID",
  "created": 1748261757,
  "search_intent": [
    {
      "query": "原始搜索query",
      "intent": "SEARCH_ALL|SEARCH_NONE|SEARCH_ALWAYS",
      "keywords": "改写后的搜索关键词"
    }
  ],
  "search_result": [
    {
      "title": "标题",
      "content": "内容摘要",
      "link": "结果链接",
      "media": "网站名称",
      "icon": "网站图标",
      "refer": "角标序号",
      "publish_date": "网站发布时间"
    }
  ]
}

工作流程

1. 确认用户已配置 API Key
2. 根据用户需求选择合适的搜索参数
3. 执行搜索并以友好格式展示结果（标题、来源、日期、链接、摘要）

错误码

错误码	说明
1701	网络搜索并发已达上限，请稍后重试
1702	系统未找到可用的搜索引擎服务
1703	搜索引擎未返回有效数据，请调整查询条件

更多信息

• 智谱 Web Search API 文档
• 产品价格