← 返回 Skills 市场
酒店比价
作者
fenbeitong-trip
· GitHub ↗
· v1.0.0
· MIT-0
103
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install hotel-price-compare
功能描述
酒店比价助手,对比携程、美团、同程、去哪儿、华住会、锦江会、飞猪等OTA平台相同酒店房型价格,给出最优推荐。Invoke when user wants to compare hotel prices across multiple OTA platforms or find the best hotel deal.
使用说明 (SKILL.md)
酒店比价助手 (fb-hotel-comparison-skill)
技能描述
酒店比价助手,实时对比携程、美团、同程、去哪儿、华住会、锦江会、飞猪等OTA平台相同酒店下所有房型产品的报价,通过智能算法给出最优推荐。
⚠️ 【重要约束】
- 必须调用各OTA平台的API获取真实价格数据
- 禁止自行编造价格信息
- 接口返回什么数据就展示什么,不要修改
- 比价结果需标明数据来源和采集时间
技能概述
基于多平台OTA API开发的酒店比价技能,支持:
- 多平台酒店价格实时抓取
- 同酒店同房型跨平台价格对比
- 智能最优推荐算法(综合考虑价格、取消政策、早餐、积分等)
- 价格趋势分析和预订建议
技能能力
核心能力
- 多平台价格抓取:同时查询携程、美团、同程、去哪儿、华住会、锦江会、飞猪等平台
- 智能比价:同酒店同房型跨平台价格对比
- 最优推荐:基于价格、政策、权益综合评分推荐最优选择
- 价格监控:支持设置价格提醒,降价通知
触发条件
-
酒店比价:当用户输入「比价/对比/哪个便宜/最优惠」+ 酒店名称时
- 调用多平台API获取价格
- 展示各平台同房型价格对比表格
- 给出最优推荐
-
展示格式示例:
🏨 桔子酒店(北京燕莎三元桥店) - 多平台比价 📅 入住:3月15日 → 退房:3月16日 🛏️ 房型:高级大床房 | 平台 | 价格 | 早餐 | 取消政策 | 会员权益 | 推荐指数 | |:---:|---:|:---:|:---|:---|:---:| | 🥇 携程 | ¥483 | 含早 | 限时取消 | 积分+早餐 | ⭐⭐⭐⭐⭐ | | 美团 | ¥526 | 无早 | 不可取消 | 无 | ⭐⭐⭐ | | 同程 | ¥498 | 含早 | 限时取消 | 积分 | ⭐⭐⭐⭐ | | 飞猪 | ¥510 | 无早 | 限时取消 | 积分 | ⭐⭐⭐⭐ | 💡 最优推荐:携程 ¥483(含早+限时取消+积分) 🔗 [立即预订-携程](https://...)
对接平台
支持OTA平台
| 平台 | 标识 | 特点 |
|---|---|---|
| 携程 | ctrip | 价格优势明显,会员权益丰富 |
| 美团 | meituan | 本地生活场景多,优惠券多 |
| 同程 | tongcheng | 微信生态,返现活动多 |
| 去哪儿 | qunar | 低价策略,学生优惠 |
| 华住会 | huazhu | 自有会员体系,积分价值高 |
| 锦江会 | jinjiang | 会员折扣力度大 |
| 飞猪 | fliggy | 阿里生态,信用住免押金 |
核心接口列表
一、查询类接口
| 接口名称 | 核心用途 | 必选参数 |
|---|---|---|
| search_hotel_prices | 多平台酒店价格查询 | hotel_name, city, check_in, check_out, platforms |
| get_price_trend | 价格趋势分析 | hotel_id, room_type, days |
| get_best_deal | 获取最优推荐 | hotel_name, city, check_in, check_out |
二、数据类接口
| 接口名称 | 核心用途 | 说明 |
|---|---|---|
| compare_platforms | 多平台对比 | 返回结构化对比数据 |
| calculate_recommendation | 计算推荐分数 | 基于多维度算法评分 |
推荐算法
评分维度
| 维度 | 权重 | 说明 |
|---|---|---|
| 价格 | 40% | 基础价格对比 |
| 取消政策 | 20% | 限时取消 > 不可取消 |
| 早餐权益 | 15% | 含早 > 无早 |
| 会员权益 | 15% | 积分、升级等 |
| 平台信誉 | 10% | 服务保障、退改政策 |
推荐指数
- ⭐⭐⭐⭐⭐ (90-100分):强烈推荐,综合最优
- ⭐⭐⭐⭐ (75-89分):推荐,某方面突出
- ⭐⭐⭐ (60-74分):一般,可考虑
- ⭐⭐ (40-59分):不推荐
- ⭐ (0-39分):强烈不推荐
响应规则
成功响应
{
"code": 0,
"msg": "success",
"data": {
"hotel_name": "桔子酒店(北京燕莎三元桥店)",
"check_in": "2026-03-15",
"check_out": "2026-03-16",
"room_type": "高级大床房",
"platforms": [
{
"platform": "携程",
"price": 483,
"breakfast": "含早",
"cancel_policy": "限时取消",
"benefits": ["积分", "早餐"],
"score": 95,
"recommendation": "⭐⭐⭐⭐⭐"
}
],
"best_deal": {
"platform": "携程",
"price": 483,
"savings": 43,
"reason": "价格最低且含早餐"
}
}
}
失败响应
{
"code": 500,
"msg": "错误信息",
"data": null
}
安全使用建议
This skill's goal (real-time cross-OTA price comparison) is plausible, but the shipped code does not implement the actual OTA queries and requests no API credentials — that's an implementation mismatch. Before installing or running: (1) ask the author how OTA authentication is provided (API keys, cookies, or scraping) and where credentials should be stored; (2) request fully implemented fetch methods or tests showing real queries to each OTA; (3) confirm dependency requirements (requests library) and run in a sandboxed environment to limit risk; (4) review any network endpoints the code will call once fetch methods are implemented; (5) do not supply high-privilege secrets (AWS, payment, or unrelated tokens) unless you validate they are necessary and handled securely. If the author cannot explain how authentication and real data fetching are handled, treat the skill as incomplete and avoid using it with sensitive credentials.
功能分析
Type: OpenClaw Skill
Name: hotel-price-compare
Version: 1.0.0
The skill bundle is a well-structured framework for a hotel price comparison tool targeting Chinese OTA platforms. The Python implementation in `hotel_comparison_api.py` and `openai_adapter.py` contains legitimate scoring logic and function-calling schemas, although the actual data-fetching methods are currently stubs (TODOs). There is no evidence of malicious intent, data exfiltration, or prompt injection in the code or the `SKILL.md` instructions.
能力评估
Purpose & Capability
The skill's description requires calling multiple OTA APIs to obtain real prices. The code includes per-OTA fetch methods, but those methods are TODO stubs returning empty lists (no real API calls implemented). The package requests network access (requests.Session) but declares no credentials or API keys even though most OTA integrations require keys/cookies or scraping credentials. Required binary is only python3 — that's plausible but insufficient for the stated capability.
Instruction Scope
SKILL.md explicitly mandates 'must call OTA APIs' and 'do not fabricate prices', but runtime code as provided will not fetch OTA prices (stubs). The instructions do not ask the agent to read unrelated files or secrets, which is good, but they assume access to OTA APIs without specifying how credentials or authentication will be provided. That mismatch grants the agent unclear discretion (e.g., how to authenticate or where to fetch data).
Install Mechanism
No install spec — lowest risk for arbitrary downloads. However, the Python scripts import the requests library but the skill does not declare or install that dependency; running may fail or implicitly rely on environment-wide packages. No external URLs, installers, or archive extraction are present.
Credentials
The skill declares no required environment variables or credentials, but real OTA API access typically requires API keys, tokens, or cookies. The absence of any credentials is disproportionate to the claimed real-time API functionality and leaves unclear how authentication will be handled — either the skill is incomplete or it expects secrets to be supplied out-of-band (not declared), which is a red flag for coherence and for potential ad-hoc secret usage.
Persistence & Privilege
The skill does not request permanent/always installation and does not modify other skills or system settings. It is user-invocable and allows model invocation (normal default). No elevated persistence or privileges are requested.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install hotel-price-compare - 安装完成后,直接呼叫该 Skill 的名称或使用
/hotel-price-compare触发 - 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
v1.0.0
- Initial release of Hotel Price Comparison skill (fb-hotel-comparison-skill).
- Supports real-time price comparison for the same hotel room type across Ctrip, Meituan, Tongcheng, Qunar, Huazhu, Jinjiang, Fliggy, and more.
- Uses OTA platform APIs for accurate, up-to-date price data; prohibits fabricating or altering price information.
- Shows side-by-side price and benefits table, best deal recommendation, and marks data source and retrieval time.
- Includes smart recommendation algorithm considering price, cancellation policy, breakfast, member benefits, and platform reputation.
- Allows users to set price alerts and view price trend analysis.
元数据
常见问题
酒店比价 是什么?
酒店比价助手,对比携程、美团、同程、去哪儿、华住会、锦江会、飞猪等OTA平台相同酒店房型价格,给出最优推荐。Invoke when user wants to compare hotel prices across multiple OTA platforms or find the best hotel deal. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 103 次。
如何安装 酒店比价?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install hotel-price-compare」即可一键安装,无需额外配置。
酒店比价 是免费的吗?
是的,酒店比价 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
酒店比价 支持哪些平台?
酒店比价 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 酒店比价?
由 fenbeitong-trip(@gaogao605)开发并维护,当前版本 v1.0.0。
推荐 Skills