← 返回 Skills 市场
yangzhichao814-cell

global-search

作者 yangzhichao814-cell · GitHub ↗ · v1.1.9 · MIT-0
cross-platform ✓ 安全检测通过
357
总下载
8
收藏
0
当前安装
27
版本数
在 OpenClaw 中安装
/install global-search
功能描述
Unleashes cutting-edge multi-source search technology that instantly synthesizes vast amounts of information across the web, delivering comprehensive and rel...
使用说明 (SKILL.md)

\r \r

Global Search\r

\r This skill performs multi-source web search through an external service and can aggregate results across multiple sources when the user explicitly asks for comprehensive coverage.\r \r

⚠️ Privacy and Security Notice\r

\r Important: This skill sends the user’s search query to an external web search service at https://clb.ciglobal.cn/web_search. Please be aware:\r \r

  • Data Transmission: Queries are transmitted to a third-party service and may be associated with your account\r
  • API Key Required: The only runtime credential is GLOBAL_SEARCH_API_KEY\r
  • Credential Storage: Store the API key securely using environment variables or your system's credential manager. Do not hardcode it in scripts\r
  • User Consent: If a query may contain sensitive, personal, confidential, or internal information, ask the user to confirm before sending it to the external service\r
  • Sensitive Query Filter: Do not send passwords, tokens, session cookies, private keys, personal identifiers, internal-only content, or other confidential data\r
  • Data Minimization: Prefer shortening, generalizing, or redacting unnecessary sensitive details before searching\r
  • Scope Limitation: Use this skill only for legitimate search purposes\r
  • Provider Verification: The provider clb.ciglobal.cn is an external third-party service. Its trustworthiness, retention policy, and privacy practices should be verified by the skill publisher before distribution; users should not assume it has been independently vetted\r
  • Account Association: The API key associates search activity with your account, so search history may be logged or retained by the provider\r \r

Credential Configuration\r

\r Required Environment Variable:\r

# Set environment variable (Linux/Mac)\r
export GLOBAL_SEARCH_API_KEY="your_api_key_here"\r
\r
# Set environment variable (Windows PowerShell)\r
$env:GLOBAL_SEARCH_API_KEY="your_api_key_here"\r
```\r
\r
Alternatively, configure the API key in your agent's credential manager or secrets storage.\r
\r
## When to Use\r
\r
Apply this skill when the user:\r
- Asks to search the web or gather information online\r
- Needs news articles or references by keyword\r
- Wants to retrieve content from diverse online sources\r
- Requires comprehensive, real-time web search with maximum coverage\r
\r
Use comprehensive search only when the user explicitly requests broad multi-source coverage. For simple lookups, prefer the minimum necessary query scope.\r
\r
## API Overview\r
\r
**Endpoint:** `POST /web_search`\r
\r
**Base URL:** `https://clb.ciglobal.cn`\r
\r
**Authentication:** Required header `X-API-Key`. Get your API key at https://clb.ciglobal.cn/apiKey/login\r
\r
**Note:** This skill sends user queries to an external web search service. Please ensure you have obtained proper authorization before using this service.\r
\r
## Request\r
\r
For comprehensive search (default behavior), the skill will use the script from overall.md to perform 4 parallel API calls automatically:\r
- **Call 1**: `search_source=baidu_search`, `mode=network` (百度新闻/资讯)\r
- **Call 2**: `search_source=google_search`, `mode=network` (谷歌新闻/资讯)\r
- **Call 3**: `search_source=baidu_search_ai`, `mode=network` (百度 AI 搜索)\r
- **Call 4**: `mode=warehouse` (Elasticsearch 索引库,忽略 search_source)\r
\r
### Headers\r
| Header | Required | Description |\r
|--------|----------|-------------|\r
| X-API-Key | Yes | API key for authentication. Get your key at https://clb.ciglobal.cn/apiKey/login |\r
| Content-Type | Yes | `application/x-www-form-urlencoded` (form data) |\r
\r
### Form Parameters\r
| Parameter | Type | Required | Default | Description |\r
|-----------|------|----------|---------|-------------|\r
| keyword | string | Yes | - | Search keyword(s),多个关键词用空格分隔 |\r
| search_source | string | No | - | Engine: `baidu_search`, `google_search`, `baidu_search_ai`. Note: Ignored when using default comprehensive search |\r
| mode | string | No | - | `network` = live crawl, `warehouse` = ES index. Note: Ignored when using default comprehensive search |\r
| page | int | No | 1 | Page number (starts from 1) |\r
\r
\r
\r
### Comprehensive Search (All Sources)\r
当用户明确要求进行全面搜索(即同时搜索所有可用来源)时,使用 `overall.md` 中的脚本进行搜索,而不是使用下面的示例代码。\r
\r
When the user explicitly wants to search across ALL available sources simultaneously (comprehensive search), you should:\r
\r
\r
**Example Implementation (Python asyncio):**\r
```python\r
import aiohttp\r
import asyncio\r
import os\r
\r
API_URL = "https://clb.ciglobal.cn/web_search"\r
# Get API key from environment variable (recommended)\r
API_KEY = os.environ.get("GLOBAL_SEARCH_API_KEY", "your_api_key_here")\r
\r
if API_KEY == "your_api_key_here":\r
    raise ValueError("Please set GLOBAL_SEARCH_API_KEY environment variable or provide your API key")\r
\r
headers = {\r
    "X-API-Key": API_KEY,\r
    "Content-Type": "application/x-www-form-urlencoded"\r
}\r
\r
SEARCH_CONFIGS = [\r
    {"name": "百度搜索", "mode": "network", "search_source": "baidu_search"},\r
    {"name": "谷歌搜索", "mode": "network", "search_source": "google_search"},\r
    {"name": "百度 AI 搜索", "mode": "network", "search_source": "baidu_search_ai"},\r
    {"name": "全库搜", "mode": "warehouse", "search_source": None}\r
]\r
\r
async def fetch_search(session, semaphore, config, keyword, page):\r
    async with semaphore:\r
        data = {\r
            "keyword": keyword,\r
            "page": page,\r
            "mode": config['mode'],\r
        }\r
        if config['search_source']:\r
            data["search_source"] = config['search_source']\r
        \r
        async with session.post(API_URL, headers=headers, data=data) as response:\r
            result = await response.json()\r
            return result.get('references', [])\r
\r
async def comprehensive_search(keyword, page=1):\r
    async with aiohttp.ClientSession() as session:\r
        semaphore = asyncio.Semaphore(5)  # Max 5 concurrent requests\r
        tasks = [fetch_search(session, semaphore, config, keyword, page) \r
                 for config in SEARCH_CONFIGS]\r
        results = await asyncio.gather(*tasks)\r
        # Flatten all references into one list\r
        all_references = [ref for refs in results for ref in refs]\r
        return all_references\r
\r
asyncio.run(comprehensive_search("人工智能"))\r
```\r
\r
### Parameter Constraints\r
- `search_source`: One of `baidu_search`, `google_search`, `baidu_search_ai`\r
- `mode`: One of `network`, `warehouse`\r
- When `mode=warehouse`, search is performed against the Elasticsearch index and ignores `search_source`\r
- When `mode=network`, use `search_source` to select Baidu, Google, or Baidu AI search\r
\r
## Response Format\r
\r
```json\r
{\r
  "code": 200,\r
  "message": "success",\r
  "references": [\r
    {\r
      "title": "Article title",\r
      "sourceAddress": "https://example.com/article",\r
      "origin": "Source name",\r
      "publishDate": "2025-03-24 12:00:00",\r
      "summary": "Article summary or snippet"\r
    }\r
  ]\r
}\r
```\r
\r
## Usage Examples\r
\r
### Example 1: Search Baidu news\r
```\r
POST https://clb.ciglobal.cn/web_search\r
Headers: X-API-Key: \x3Cyour_api_key>, Content-Type: application/x-www-form-urlencoded\r
Body (form): keyword=人工智能&search_source=baidu_search&mode=network&page=1\r
```\r
\r
### Example 2: Search Google news\r
```\r
POST https://clb.ciglobal.cn/web_search\r
Headers: X-API-Key: \x3Cyour_api_key>, Content-Type: application/x-www-form-urlencoded\r
Body (form): keyword=AI&search_source=google_search&mode=network&page=1\r
```\r
\r
### Example 3: Search warehouse (ES index)\r
```\r
POST https://clb.ciglobal.cn/web_search\r
Headers: X-API-Key: \x3Cyour_api_key>, Content-Type: application/x-www-form-urlencoded\r
Body (form): keyword=机器学习&mode=warehouse&page=1\r
```\r
\r
### Example 4: cURL\r
```bash\r
curl -X POST "https://clb.ciglobal.cn/web_search" \\r
  -H "X-API-Key: \x3Cyour_api_key>" \\r
  -H "Content-Type: application/x-www-form-urlencoded" \\r
  -d "keyword=科技新闻&search_source=baidu_search&mode=network&page=1"\r
```\r
\r
## Error Codes\r
\r
| Code | Message | Cause |\r
|------|---------|-------|\r
| 401 | X-API-Key参数缺失 | Missing API key header |\r
| 401 | ApiKey无效 | Invalid API key |\r
| 400 | search_source参数错误 | Invalid search_source value |\r
| 400 | mode参数错误 | Invalid mode value |\r
| 400 | page参数错误 | Invalid page (non-integer or 0) |\r
\r
## Integration Steps\r
\r
1. **Verify Provider:** Before publishing or using this skill, verify `clb.ciglobal.cn`’s trust model, privacy policy, data retention practices, and ownership information\r
2. **Obtain API Key:** Visit https://clb.ciglobal.cn/apiKey/login to get your API key\r
3. **Configure Credentials:** Set the `GLOBAL_SEARCH_API_KEY` environment variable with your API key\r
   - Linux/Mac: `export GLOBAL_SEARCH_API_KEY="your_api_key_here"`\r
   - Windows: `$env:GLOBAL_SEARCH_API_KEY="your_api_key_here"`\r
4. **API Base URL:** `https://clb.ciglobal.cn`\r
5. **Comprehensive Search:** Only perform comprehensive search across all sources when the user explicitly requests broad multi-source coverage\r
6. **Sensitive Query Check:** If the user’s query includes sensitive, personal, confidential, or internal data, refuse to transmit it and ask for a redacted or generalized query instead\r
7. **Make Request:** Call `POST /web_search` with form-encoded parameters and include `X-API-Key` header\r
8. **Parse Response:** Parse `references` from the response and use `title`, `sourceAddress`, `summary` as needed\r
安全使用建议
This skill forwards user queries to a third-party service (https://clb.ciglobal.cn) and requires an API key (GLOBAL_SEARCH_API_KEY). Before installing, verify the provider (privacy policy, retention, who runs clb.ciglobal.cn) because search queries may be logged and associated with your account. Never send passwords, tokens, or other secrets in queries; ask users for consent before searching sensitive or personal information. Store the API key securely (env var or credential manager) and rotate/revoke it if you stop using the skill or suspect misuse. If you need an audited/trusted provider, consider using a known vendor or an internal enterprise search service instead.
功能分析
Type: OpenClaw Skill Name: global-search Version: 1.1.9 The 'global-search' skill is a legitimate integration for a multi-source web search API hosted at clb.ciglobal.cn. The provided Python scripts (overall.md) and documentation (SKILL.md, DEMO.md) use standard libraries like aiohttp and requests to perform searches across Baidu and Google. The skill includes significant security and privacy warnings, advising users against sending sensitive data and providing clear instructions on secure credential management via environment variables. No evidence of data exfiltration, malicious execution, or prompt injection was found.
能力标签
requires-sensitive-credentials
能力评估
Purpose & Capability
The skill claims to perform multi-source web search and its instructions and demo scripts show POST calls to an external search API; the single required environment variable (GLOBAL_SEARCH_API_KEY) is appropriate for that function.
Instruction Scope
SKILL.md and overall.md explicitly instruct the agent to send search queries to https://clb.ciglobal.cn/web_search, run up to four parallel source queries for comprehensive searches, and includes privacy guidance (ask for consent, avoid sending secrets). The instructions do not ask the agent to read unrelated local files, other credentials, or transmit data to unexpected endpoints.
Install Mechanism
This is an instruction-only skill with no install spec and no code files that execute locally (only example/demo scripts). That minimizes disk-writing risk and there are no third-party downloads or packaged installs to evaluate.
Credentials
Only GLOBAL_SEARCH_API_KEY is requested and that variable is used consistently in the scripts and documentation. No unrelated secrets or system config paths are required.
Persistence & Privilege
The skill is not forced-always (always: false), does not request system-level configuration changes, and does not modify other skills. It operates by making network calls when invoked.
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 global-search
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /global-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
v1.1.9
- Added homepage and publisher metadata fields to the skill description. - No changes to the functional code or documentation content. - Version bump for metadata update only.
v1.1.8
Global-search version 1.1.8 — security notice update - Privacy and security section updated with a new warning to avoid sending passwords, tokens, session cookies, keys, personal identifiers, internal content, and other confidential data as queries - Clarified that the external provider's trustworthiness and policies must be verified by the publisher; users are advised not to assume independent vetting - No functional, API, or feature changes to the skill itself
v1.1.7
global-search v1.1.7 - Clarified that multi-source (comprehensive) search is performed only when explicitly requested by the user. - Updated privacy and consent instructions: recommend confirming with the user and minimizing sensitive query details. - Emphasized credential management and trust evaluation for the third-party provider. - Adjusted usage guidance to favor minimum necessary search scope for simple queries. - Minor rewording and formatting improvements for conciseness and clarity.
v1.1.6
- Expanded the Privacy and Security Notice to include provider trustworthiness verification and a warning that search history is linked to your API key/account. - Clarified that users should review the third-party provider's privacy policy before use. - Added details that the provider may log and store all queries associated with the API key. - No functional/API changes.
v1.1.5
- Added required_env_vars (GLOBAL_SEARCH_API_KEY) to skill metadata for clearer environment configuration. - No changes to logic, endpoints, or usage—documentation and metadata only.
v1.1.4
**Summary:** Added explicit privacy, security, and credential configuration guidance. - Introduced a "Privacy and Security Notice" section highlighting data transmission risks, consent, and use limitations. - Added instructions for using environment variables (`GLOBAL_SEARCH_API_KEY`) to securely store the API key. - Updated example code to obtain the API key from the environment variable rather than hardcoding. - Clarified best practices for credential storage and user consent when using third-party search services. - General content and formatting improvements for clarity.
v1.1.3
- Added an explicit note that user queries are sent to an external web search service, and users should ensure they have proper authorization before using it. - No other changes; functionality and usage remain the same.
v1.1.2
- Now requires authentication using the X-API-Key header for all API requests. - Documentation updated to include details for obtaining and using the required API key. - Added examples with the X-API-Key header for all request samples. - New error codes for missing or invalid API keys are now documented.
v1.1.1
- Initial release of the global-search skill, offering multi-source web search with intelligent aggregation. - Supports comprehensive, parallel searches across Baidu, Google, Baidu AI, and an Elasticsearch index with a single query. - No authentication required; free service. - Default behavior executes four parallel API calls to maximize coverage and relevance. - Detailed API documentation, usage examples, and error handling provided. - Ideal for retrieving news, articles, or references from diverse online sources in real time.
v1.0.17
- No file changes detected in this version. - No updates, bug fixes, or new features introduced. - Functionality and documentation remain unchanged from previous version.
v1.0.16
- Enhanced description for greater clarity on multi-source parallel search and intelligent aggregation. - Emphasized instant, comprehensive results from diverse online repositories. - Updated guidance to highlight maximum coverage and superior search efficiency. - Refined skill’s usage scenarios and capabilities for better audience understanding. - No changes to API endpoints or technical integration.
v1.0.15
- API endpoint updated from http://101.245.108.220:9004 to https://clb.ciglobal.cn, now using secure HTTPS. - Skill name changed from "whole-network-search" to "global-search" and branding uniformly updated. - Removed privacy and security warnings related to unencrypted HTTP and third-party data exposure. - All example requests, code samples, and documentation now use the new base URL. - No changes to parameters or functionality; search features and usage remain the same.
v1.0.14
- Enhanced the privacy and security warning with a much clearer, stronger alert about the risks of using this skill, including warnings about data interception, third-party access, and potential data storage. - Explicitly advises users to avoid searching any sensitive or personal information and consider safer alternatives. - No functional/API changes; only documentation updated for increased transparency and risk awareness.
v1.0.13
- Added a privacy and security warning to alert users that search queries are sent over unencrypted HTTP to a third-party server. - Emphasized the potential risks of transmitting sensitive or personal information due to lack of encryption and third-party IP hosting. - No changes to functionality or API—documentation update only.
v1.0.12
- No file changes detected in this version. - Functionality and documentation remain unchanged from the previous release.
v1.0.11
- No code or documentation changes detected in this release. - Version 1.0.11 is unchanged from the previous release.
v1.0.10
- Removed mention of API key configuration; clarified that no API key is needed (free service). - Updated integration steps to reflect the removal of API key requirement. - No other functional or API changes; all usage and behavior remains the same.
v1.0.9
- Authentication is no longer required; the API is now a free service. - The X-Appbuilder-Authorization header and related documentation were removed. - API usage examples and code snippets were updated to exclude the API key and authentication header. - Error codes related to missing or invalid authentication were removed.
v1.0.8
No changes detected in this version. - No file or documentation changes were made in version 1.0.8.
v1.0.7
**Comprehensive search across all sources is now the default behavior.** - By default, the skill performs simultaneous searches across Baidu, Google, and the Elasticsearch index for the most complete results. - Description updated to clearly state comprehensive search is the default. - API instructions revised to highlight parallel calls and clarify default parameters. - Manual selection of search source or mode can override the default if desired.
元数据
Slug global-search
版本 1.1.9
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 27
常见问题

global-search 是什么?

Unleashes cutting-edge multi-source search technology that instantly synthesizes vast amounts of information across the web, delivering comprehensive and rel... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 357 次。

如何安装 global-search?

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

global-search 是免费的吗?

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

global-search 支持哪些平台?

global-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 global-search?

由 yangzhichao814-cell(@yangzhichao814-cell)开发并维护,当前版本 v1.1.9。

💬 留言讨论