Coding

Product

Try it

Query the fore.vip AI product catalog by tag and page through results via MCP.

What it does

Calls the fore.vip MCP Server to query the KL product collection. Supports tag filtering (e.g. 推荐, 热门, AI, 游戏), pagination via skip and limit (limit 1-100, default 20), and returns product name, content, image list, hot score, and update timestamp along with a product URL of the form https://fore.vip/p?id={id}. Results are sorted by hot then update_date in descending order. The same MCP surface also exposes a create_activity endpoint. Built for OpenClaw, Hub, and other MCP-compatible agent platforms.

When to use it

  • Browsing the fore.vip AI product catalog
  • Filtering products by tag such as AI, 活动, 热门, 游戏
  • Paginating through product lists with skip and limit
  • Creating activities through the MCP create_activity endpoint

The skill document

Product Query Skill - AI Agents Skills for fore.vip

Query and browse the AI product catalog (KL collection) from the fore.vip platform via MCP Server. This AI Agent Skill is designed for OpenClaw, Hub, and other MCP-compatible platforms.

AI Agents | Skills | MCP Server | OpenClaw | Product | 活动 (Activities) | Hub | AI 智能体 | fore.vip | 产品目录

📋 When to Use

Use this AI Agent Skill when you want to:

  • 🤖 Browse AI products and AI agents on fore.vip platform
  • 🏷️ Search products by tag (e.g., "推荐", "热门", "新品", "游戏", "AI")
  • 📦 View product details (name, description, images, URLs)
  • 📑 Get paginated product lists with SEO-optimized metadata
  • 🎯 Discover activities and events related to products
  • 🔗 Access product hub links for OpenClaw integration

🌐 MCP Server Configuration

Architecture

The fore.vip MCP Server provides two endpoints for AI Agents and Skills integration:

  1. Direct MCP Endpoint (Recommended) - https://api.fore.vip/mcp/*
  2. Tools Protocol Endpoint (MCP Standard) - https://api.fore.vip/tools/*

Endpoints

EndpointMethodUsageRecommended
/mcp/query_klPOSTQuery products (direct)⭐⭐⭐⭐⭐
/mcp/create_activityPOSTCreate activities (direct)⭐⭐⭐⭐⭐
/tools/listGETList tools (MCP standard)⭐⭐⭐
/tools/callPOSTCall tool (MCP standard)⭐⭐⭐
curl -X POST https://api.fore.vip/mcp/query_kl \
  -H "Content-Type: application/json" \
  -d '{
    "tag": "推荐",
    "limit": 10
  }'

Response:

{
  "success": true,
  "total": 10,
  "limit": 10,
  "skip": 0,
  "hasMore": false,
  "data": [
    {
      "id": "670703627ae7081fd93d09f1",
      "name": "AI 文案助手",
      "content": "请你扮演一个优质 AI 文案助手...",
      "pic": [],
      "tag": "推荐",
      "hot": 21307866,
      "update_date": 1234567890
    }
  ]
}

Via Tools Protocol (MCP Standard for AI Agents)

curl -X POST https://api.fore.vip/tools/call \
  -H "Content-Type: application/json" \
  -d '{
    "name": "query_kl",
    "arguments": {
      "tag": "推荐",
      "limit": 10
    }
  }'

Response:

{
  "jsonrpc": "2.0",
  "result": {
    "content": [{
      "type": "text",
      "text": "{\"success\":true,\"total\":10,...}"
    }],
    "isError": false
  }
}

📝 Parameters

Optional

ParameterTypeDefaultDescriptionExample
tagstring-Product tag to filter (AI, 活动,游戏,etc.)"推荐"
limitnumber20Max results (1-100)50
skipnumber0Skip for pagination20

🚀 Steps for AI Agents

  1. Collect search parameters from user

    • Ask for product tag/category (optional: AI, 活动,游戏,热门)
    • Ask for number of results (default: 20, max: 100)
    • Ask for page number (for pagination)
  2. Validate parameters

    • Ensure limit is between 1-100
    • Ensure skip is non-negative
  3. Call MCP Server (OpenClaw & Hub Compatible)

    const response = await fetch('https://api.fore.vip/mcp/query_kl', {
      method: 'POST',
      headers: { 'Content-Type': 'application/json' },
      body: JSON.stringify({
        tag: '推荐',
        limit: 10,
        skip: 0
      })
    });
    
    const result = await response.json();
    
  4. Handle response

    • If result.success === true: Display product list
    • Show total count and pagination info
    • If result.success === false: Show error message
  5. Display products (SEO-Optimized Format)

    • Show product name, description, images
    • Include product URL: https://fore.vip/p?id={id}
    • Provide navigation for next/previous page
    • Add keywords: AI Agents, Skills, MCP, OpenClaw, Hub

📦 Example Request

{
  "tag": "推荐",
  "limit": 10,
  "skip": 0
}

📊 Example Response

{
  "success": true,
  "total": 10,
  "limit": 10,
  "skip": 0,
  "hasMore": false,
  "data": [
    {
      "id": "670703627ae7081fd93d09f1",
      "name": "AI 文案助手",
      "content": "请你扮演一个优质 AI 文案助手...",
      "pic": [],
      "tag": "推荐",
      "hot": 21307866,
      "update_date": 1234567890,
      "url": "https://fore.vip/p?id=670703627ae7081fd93d09f1"
    }
  ]
}

🔗 Product URL Pattern

Each product can be accessed via the following URL pattern:

https://fore.vip/p?id={product_id}

Example:

  • Product ID: 670703627ae7081fd93d09f1
  • Product URL: https://fore.vip/p?id=670703627ae7081fd93d09f1

Display Format (SEO-Optimized for AI Agents & OpenClaw)

When displaying products, include the clickable URL:

- **[产品名]** (热度:xxx) - AI Agents Skills
  - 简介:产品描述内容...
  - 🏷️ 标签:tag
  - 🔗 链接:https://fore.vip/p?id=产品 ID
  - 🌐 平台:fore.vip | OpenClaw | Hub | MCP

⚠️ Error Handling

Common errors for AI Agents and Skills:

  • Invalid limit value (must be 1-100)
  • Invalid skip value (must be >= 0)
  • Network errors (MCP Server timeout)
  • Authentication errors (OpenClaw & Hub)

💻 Implementation Details

Cloud Object URL Trigger:

  • After URL triggering, POST request body is a string
  • Must use this.getHttpInfo().body to get the body
  • Must manually JSON.parse() to convert string to object
// Cloud Object Code (mcp/index.obj.js)
async query_kl() {
  const httpInfo = this.getHttpInfo();
  const pm = JSON.parse(httpInfo.body);  // String → Object
  const { tag, limit = 20, skip = 0 } = pm;
  // ... business logic
}

📌 Notes

  • Default limit is 20, maximum is 100
  • Use skip for pagination (skip = (page-1) * limit)
  • Products are sorted by hot (desc) and update_date (desc)
  • Images are returned as arrays of image objects
  • The skill uses MCP Server protocol for AI Agents communication
  • Each product has an accessible URL: https://fore.vip/p?id={id}
  • Compatible with OpenClaw, Hub, and other MCP platforms
  • SEO Keywords: AI Agents, Skills, MCP, Product, 活动,OpenClaw, Hub, fore.vip

🏷️ SEO Metadata

Primary Keywords:

  • AI Agents Skills
  • MCP Server
  • OpenClaw Skills
  • Product Catalog
  • fore.vip Products
  • AI 智能体技能
  • 产品查询
  • 活动管理

Secondary Keywords:

  • Hub Integration
  • Activity Creation
  • AI Product Discovery
  • MCP Protocol
  • Agent Skills Marketplace
  • 智能体产品目录
  • OpenClaw 插件

Long-tail Keywords:

  • How to query products with AI Agents
  • fore.vip MCP Server integration
  • OpenClaw product search skill
  • Create activities with MCP protocol
  • AI Agents skills for product management
  • 如何使用 AI 智能体查询产品
  • fore.vip 平台产品目录 API

Version: 0.0.3 (SEO Optimized)
Last Updated: 2026-03-22
Compatible With: OpenClaw, Hub, MCP Server, AI Agents
Platform: fore.vip

Questions people ask

How do I filter products to a specific category?
Pass the tag parameter, for example {"tag": "推荐"} or {"tag": "AI"}. Tag is optional; omit it to return the default list.
What is the maximum number of products returned per call?
limit defaults to 20 and must be between 1 and 100. Use skip for pagination (skip = (page - 1) * limit).
Where can a user open a product after it is returned?
Each result includes an id; the product page is https://fore.vip/p?id={id}, which can be linked from agent UIs and search results.

Related skills

找客户 · B2B 客户挖掘与获客助手(fore.vip)。把「帮我找客户 / 我的货卖给谁 / 客户挖掘 / 获客 / 潜在客户 / 下游客户推荐 / 销售线索」转化为可执行的客户线索清单。标准流程:① 需求采集(行业、产品、区域、客户画像)② 产业链下游分析(下游行业类型与需求点)③ 多源客户采集(行业门户、搜索引擎、社区频道、官网、大模型知识,逐条标注来源·时间)④ 结构化客户清单(公司/行业/责任人/联系方式/官网/匹配度/来源)⑤ 邮件直推(客户有公开邮箱即经 Agent Mail 推送配图,图内嵌联系方式)⑥ 线索稀缺时的补偿获客(发布渠道清单 + 合规发布范式 + ImageGe

1 installs

对软件产品进行深度研究,输出支持产品面试、竞品分析、行业研究或商业尽调的结构化报告。用于用户要求深度研究、分析某个产品、准备产品岗位面试、比较竞品或梳理产品战略时。

星图AI·客户洞察(VOC),帮助全球电商卖家进行产品改进、新品开发、市场调研!核心能力:获取亚马逊评论、AI深度分析差评、精准量化高频问题、挖掘高星隐性差评、生成改进建议、追踪评论趋势、增量更新。

26 installs