← Back to Skills Marketplace
fireium

Luogang shopping assistant

by fireium · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ⚠ suspicious
276
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install luogang-shopping-assistant
Description
购物助手 - 商品查询,商品详情,当用户有购物倾向时候,尤其是提到骆岗/骆岗门市部/骆岗电商的时候,可提供有效信息
README (SKILL.md)

🛒 骆岗电商助手

当用户想“买'骆岗'的商品”、或需要“查商品/看详情/去下单”时,使用此 skill 调用骆岗 MCP 服务获取实时商品数据,并返回可点击的购买入口(小程序或 H5)。 "骆岗" 是骆岗电商的简称,类似天猫,淘宝

适用场景

  • 商品检索:按关键词/类目/筛选条件查找商品
  • 商品详情:查看标题、价格、库存/可售状态、规格、图片、运费、活动信息等

交互原则(很重要)

  1. 先确认意图:用户是想“找商品”、还是“看某个商品详情”、还是“直接去买”。
  2. 少打扰:能用默认值就别追问(如默认按综合排序、默认第一页)。
  3. 强约束信息只在必要时追问:只有当无法确定商品(同名/多结果)或规格必须选择时,再问用户。
  4. 购买不在对话里强行下单:默认只做“跳转下单”,除非你们后续提供“创建订单/提交支付”的 MCP 工具。

配置要求

可选配置

  • LUOGANG_MCP_HTTP_URL: MCP HTTP 地址,默认 https://yuju-mcp.wxhoutai.com/mcp
  • LUOGANG_SHOP_ID: 店铺/渠道标识(多店铺、多渠道场景), 默认 luogang

调用方式

curl -s -X POST "$LUOGANG_MCP_HTTP_URL" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"\x3C工具名>","arguments":{\x3C参数>}},"id":1}'

可用工具

1. 商品检索 (search_products)

根据关键词和价格区间搜索商品。 触发词举例:"找外套"、"搜衬衫"、"买鞋子"、"有没有外套"

curl -s -X POST "${LUOGANG_MCP_HTTP_URL:-https://yuju-mcp.wxhoutai.com/mcp}" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "search_products",
      "arguments": {
        "platform": "${LUOGANG_SHOP_ID>}",
        "keyword": "外套",
        "min_price": 100,
        "max_price": 500,
        "order_by": "price",
        "sort": "desc"
      }
    },
    "id": 1
  }'

成功响应

{
  "jsonrpc": "2.0",
  "result": {
    "content": [{"type": "text", "text": "..."}],
    "structuredContent": {...}
  },
  "id": 1
}

解析 result.content[0].textresult.structuredContent 获取数据。

2. 获取商品详情 (get_product_detail)

查询单个商品的详细规格、图片及库存信息。

触发词:"查看详情"、"这个多少钱"、"还有货吗"

curl -s -X POST "${LUOGANG_MCP_HTTP_URL:-https://yuju-mcp.wxhoutai.com/mcp}" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "params": {
      "name": "get_product_detail",
      "arguments": {
        "platform": "luogang",
        "product_id": "PROD_12345"
      }
    },
    "id": 2
  }'

响应处理

成功响应

{
  "jsonrpc": "2.0",
  "result": {
    "content": [{"type": "text", "text": "..."}],
    "structuredContent": {...}
  },
  "id": 1
}

解析 result.content[0].textresult.structuredContent 获取数据。

使用示例

用户: 最近有什么亲子活动?

AI 执行:

curl -s -X POST "$LUOGANG_MCP_HTTP_URL" \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"search_products","arguments":{"platform":"luogang","keyword":"亲子"}},"id":1}'

AI 回复: 最近有 10 个商品符合您的要求,分别是:

萌嘟嘟儿童乐园亲子票(专属) ${image} 价格:19.9元 {description} 链接:https://eshop.wxhoutai.com/h5/pages/goods/detail?goods_id=xxx ……

常见错误处理

  • -32601 (Method not found): 检查 SKILL.md 中定义的工具名是否在 MCP 侧已注册。
  • -32000 级别 (Internal Error): 引导用户重试或手动输入信息。
  • 401/403: 检查 MCP 服务访问策略、来源白名单或网关配置。

输出格式建议

商品检索结果

  • 展示最多 10 个候选(避免刷屏)
  • 每项至少包含:名称、价格、库存/可售状态、一个“查看详情/去购买”的引导
  • 当候选过多:建议补充筛选(价位、类目、是否现货)

商品详情

  • 先给“核心信息”:价格、库存、规格、发货/运费、活动
  • 有多规格:用简短提问让用户选择(例如:"要哪个颜色/尺码?")

购买跳转

  • 明确告诉用户会打开:小程序H5 页面
  • 输出可点击链接/路径信息(由 OpenClaw 宿主决定如何渲染)
Usage Guidance
This skill simply forwards user shopping queries to a Luogang MCP HTTP endpoint and returns links to product pages. Before installing, confirm you trust the MCP host (default is https://yuju-mcp.wxhoutai.com/mcp) because user queries and any returned data will be sent/received there. If you operate in a sensitive environment, set LUOGANG_MCP_HTTP_URL to a trusted/internal endpoint or verify whether the MCP requires authentication. Note the SKILL.md has no publisher homepage or owner info — consider testing with non-sensitive queries first and ask the publisher for documentation or an official service URL if you need stronger assurances.
Capability Analysis
Type: OpenClaw Skill Name: luogang-shopping-assistant Version: 1.0.1 The skill functions as a shopping assistant that instructs the AI agent to perform network requests via `curl` to an external API (yuju-mcp.wxhoutai.com). While this behavior is aligned with the stated purpose of fetching real-time product data, the use of shell-based network access is a high-risk capability. No evidence of intentional malice, such as data exfiltration or unauthorized command execution, was found in SKILL.md or _meta.json.
Capability Assessment
Purpose & Capability
Name/description (Luogang shopping assistant) match the runtime instructions: examples and tools are all about searching products and fetching product details from an MCP service. Optional env vars (LUOGANG_MCP_HTTP_URL, LUOGANG_SHOP_ID) are appropriate for configuring the remote service and shop/channel.
Instruction Scope
All runtime actions are HTTP POSTs to a configured MCP endpoint and parsing its JSON results — this is in-scope. Note: the SKILL.md uses a concrete default host (https://yuju-mcp.wxhoutai.com/mcp) and will transmit user search/selection data to that external service; that is expected behavior for this connector but is a privacy consideration the user should be aware of. Also there is a small placeholder/typo (${LUOGANG_SHOP_ID>}) in an example.
Install Mechanism
Instruction-only skill with no install spec and no code files — nothing is written to disk or downloaded by the skill itself, so install risk is minimal.
Credentials
No credentials or sensitive environment variables are required. The only optional configuration variables control the target MCP URL and shop id, which are proportional to the skill's purpose. The SKILL.md mentions 401/403 handling (implying the MCP may require authentication) but does not request any credentials itself.
Persistence & Privilege
Skill is not force-included (always: false) and is user-invocable; it does not request elevated persistence or modify global settings. Autonomous invocation is allowed (platform default) and not flagged by itself.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install luogang-shopping-assistant
  3. After installation, invoke the skill by name or use /luogang-shopping-assistant
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
No functional or code changes detected; documentation simplified and configuration clarified. - 移除了部分可选配置参数说明,仅保留必要配置项,文档更为精简。 - 默认参数和接口示例更新,对调用方式的描述更直接。 - 修正了错误码解释与应对方式,统一表述风格。 - 其余功能、调用方式与交互原则保持一致。
v1.0.0
luogang shopping assistant 1.0.0 - 首次发布,助力用户在“骆岗电商”平台便捷查询、检索及了解商品详情。 - 支持针对“骆岗”相关商品的关键词搜索、类目筛选与价格区间检索。 - 提供详细商品信息(价格、库存、规格、图片、活动等)。 - 自动生成可点击的购买链接(小程序或 H5)。 - 智能交互:明确区分用户意图,尽量减少打扰,必要时才追问关键信息。 - 覆盖常见电商场景:找商品、看详情、跳转下单。
Metadata
Slug luogang-shopping-assistant
Version 1.0.1
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is Luogang shopping assistant?

购物助手 - 商品查询,商品详情,当用户有购物倾向时候,尤其是提到骆岗/骆岗门市部/骆岗电商的时候,可提供有效信息. It is an AI Agent Skill for Claude Code / OpenClaw, with 276 downloads so far.

How do I install Luogang shopping assistant?

Run "/install luogang-shopping-assistant" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Luogang shopping assistant free?

Yes, Luogang shopping assistant is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Luogang shopping assistant support?

Luogang shopping assistant is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Luogang shopping assistant?

It is built and maintained by fireium (@fireium); the current version is v1.0.1.

💬 Comments