← Back to Skills Marketplace
平替旅行家
by
hello_hang
· GitHub ↗
· v1.0.1
· MIT-0
100
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install flyai-destination-substitute
Description
平替旅行家——想去A太贵？帮你找平替！AI理解你向往某个目的地的"核心体验元素"，然后在更近/更便宜的目的地中找到能提供相似体验的"平替方案"。当用户提到"想去XX但太贵"、"XX平替"、"找平替"、"去不起XX"、"类似XX的地方"、"XX太远了"、"XX签证难"时使用。
README (SKILL.md)
平替旅行家 — 想去 A 太贵？帮你找平替
你是一个能够自主学习、持续成长的平替旅行规划大师。传统逻辑是「硬攒钱去梦想目的地」，你反过来——「理解梦想本质 → 找能满足它的平替方案」。
核心定位
FlyAI 能力
完整命令参考见 reference 目录
本技能主要使用：search-poi、search-hotel、search-flight、ai-search
体验本质专家：
- 🎯 洞察本质：用户说"想去瑞士"，你理解他想要的是「雪山草甸 + 精致小镇 + 田园牧歌」
- 🔄 平替匹配：在预算/时间/签证可行的目的地中，找到能满足这些体验元素的"平替"
- 📊 差异透明：诚实标注"平替了什么"和"差在哪"，让用户知情选择
- 🧬 记忆成长：记住用户的偏好和历史方案，持续提供个性化推荐
Memory 系统
作为一个能持续成长的智能助手，我会记住你的风格和偏好。
核心要点：
- 启动时读取：除非用户说"忽略偏好/换个风格"
- 有记录：直接用已保存的偏好开始对话
- 无记录：首次用户，收集基本信息
- 实时更新：用户提到出发城市、偏好、约束条件时更新 Memory
工具说明
用户画像读取（双模式）
启动时读取用户历史偏好，减少重复询问。
优先：search_memory(query="用户旅行画像", category="user_hobby", keywords="flyai")
降级：read_file(file_path="~/.flyai/user-profile.md")
工作流程
详细步骤见 reference/workflow.md
核心阶段：
- 收集用户信息 - 梦想目的地/去不了原因/出发城市
- 拆解体验元素 - 理解梦想本质（核心创新点）
- 搜索平替目的地 - 调用 FlyAI 匹配满足体验元素的目的地
- 计算匹配度 - 体验元素匹配评分
- 输出平替方案 - 含费用对比和差距说明
常见梦想目的地的平替参考
| 梦想目的地 | 国内平替 | 海外低成本平替 |
|---|---|---|
| 瑞士 🇨🇭 | 甘南、阿尔山、新疆喀纳斯 | 清迈拜县、斯洛文尼亚 |
| 马尔代夫 🏝 | 西沙群岛、涠洲岛 | 仙本那、长滩岛、甲米 |
| 圣托里尼 🇬🇷 | 青海茶卡、舟山嵊泗 | 蓝梦岛、涛岛 |
| 冰岛 🇮🇸 | 黑龙江漠河、西藏纳木错 | 挪威罗弗敦、法罗群岛 |
| 北欧 🇳🇴 | 长白山、九寨沟 | 芬兰拉普兰 |
| 新西兰 🇳🇿 | 川西、西藏林芝 | 斯里兰卡、塔斯马尼亚 |
| 土耳其·卡帕 🎈 | 敦煌、张掖 | 缅甸蒲甘 |
| 普罗旺斯 🌸 | 新疆伊犁、云南罗平 | 北海道富良野 |
现实约束与失败处理
| 情况 | 处理方式 |
|---|---|
| 梦想目的地太独特无法平替 | 诚实告知"XX体验很难平替"，推荐最接近的方案并标注差距 |
| 用户不接受"平替"概念 | 转为正面推荐："基于你喜欢XX的偏好，推荐这些目的地" |
| 平替目的地的数据少 | 用 keyword-search 补充，标注"景点信息有限" |
| 季节不匹配 | 明确标注"该平替X月去最佳，你的时间去可能体验打折" |
| 预算仍然不够 | 给出"如果多加¥X就能解锁XX"的升级提示 |
| SSL 证书验证失败 | 确保命令前加 NODE_TLS_REJECT_UNAUTHORIZED=0 |
自主学习机制
示例对话
用户偏好保存（双模式）
发现新偏好时提示保存。详见 reference/user-profile-storage.md
保存流程：发现偏好 → 提示确认 → Qoder用update_memory / 非Qoder更新本地文件
Usage Guidance
This skill appears to do what it says (find cheaper substitute destinations) but includes insecure and inconsistent runtime instructions you should consider before installing or running:
- Ask the author to explain and fix the conflicting CLI package names (@fly-ai vs @anthropic-ai). Don’t run installs until that is clarified.
- Do NOT accept or use NODE_TLS_REJECT_UNAUTHORIZED=0; it disables TLS certificate checks and can enable MITM attacks and silent credential leakage. Request a version of the skill that does proper TLS verification.
- Avoid running global npm installs (npm install -g) with sudo on production/shared machines. If you must try this skill, run the CLI in an isolated environment (container/VM) and inspect the package source before installing.
- Be aware the skill will read/write a local profile file (~/.flyai/user-profile.md) and may persist personal preferences; if you care about privacy, inspect the file format and limit what is stored.
If the author can (1) remove the TLS-bypass instruction, (2) clarify which CLI package to install and why, and (3) provide either a vetted install method or an option to operate without installing global packages, the safety and coherence concerns would be greatly reduced.
Capability Analysis
Type: OpenClaw Skill
Name: flyai-destination-substitute
Version: 1.0.1
The skill bundle instructs the AI agent to perform a global system installation (npm install -g @fly-ai/flyai-cli) and explicitly directs it to bypass SSL certificate validation (NODE_TLS_REJECT_UNAUTHORIZED=0) when executing CLI commands, as seen in workflow.md. It also implements a persistent storage mechanism that reads and writes user data to the home directory (~/.flyai/user-profile.md). While these behaviors appear aligned with the travel planning purpose, they introduce significant security risks, including potential Man-in-the-Middle (MITM) vulnerabilities and supply chain risks through automated global package installation.
Capability Assessment
Purpose & Capability
The skill's name/description (finding cheaper substitute destinations) aligns with the commands and search tools it calls (search-poi, search-hotel, search-flight, ai-search). Reading and saving a user travel profile is reasonable for personalization. However, the workflow requires installing a third-party CLI at runtime (npm install -g @fly-ai/flyai-cli@latest) and another reference to npx @anthropic-ai/flyai-cli@latest appears in tools.md — the two different package namespaces are inconsistent and unexplained.
Instruction Scope
Instructions tell the agent to read/write a local file (~/.flyai/user-profile.md) (expected for user profile storage), and to use platform memory tools if available (search_memory/update_memory) — these are within purpose. Major concern: multiple places instruct setting NODE_TLS_REJECT_UNAUTHORIZED=0 to bypass TLS certificate validation. Explicit guidance to disable TLS validation is insecure and disproportionate to the skill's travel-planning purpose.
Install Mechanism
Although the skill is instruction-only (no install spec), the SKILL.md tells the agent to run npm install -g @fly-ai/flyai-cli@latest (registry: npmjs). Installing a global npm package at runtime has moderate risk (arbitrary code from npm). Additionally, tools.md refers to npx @anthropic-ai/flyai-cli@latest — conflicting package names create ambiguity about which package will be installed/called. The package sources are public registries (not shorteners), but the inconsistency and runtime global install increase risk.
Credentials
The skill declares no required environment variables, yet repeatedly instructs using NODE_TLS_REJECT_UNAUTHORIZED=0 when running commands, effectively asking the environment to ignore TLS validation. It also suggests using sudo for npm installs if permissions are lacking. Asking for these environment modifications (TLS bypass, elevated privileges) is not proportional to a travel-recommendation assistant and exposes network and host risks. Access to ~/.flyai is reasonable for profile storage but persistent local writes may contain sensitive personal data.
Persistence & Privilege
always is false and the skill's persistence is limited to its own user-profile storage (Qoder Memory or ~/.flyai/user-profile.md). Requesting to save preferences is coherent with personalization. No evidence the skill modifies other skills or system-wide settings. Autonomous invocation is allowed (platform default) but not combined with other high privileges here.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install flyai-destination-substitute - After installation, invoke the skill by name or use
/flyai-destination-substitute - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
v1.0.1 adds detailed reference documentation and modularizes internal logic.
- Added 13 reference documents covering AI搜索、工具用法、搜索命令、用户画像、工作流示例等
- 主文件简化，具体用法/流程/机制统一外链至 reference 目录，结构更加清晰
- 新增“用户画像读取/保存”与“自主学习机制”说明，提升个性化推荐与学习能力
- 便于维护和扩展，后续文档支持快速补充/查阅
- 实际功能与核心定位保持不变，仅优化结构和说明文档
v1.0.0
flyai-destination-substitute 1.0.0 Changelog
- Initial release of "平替旅行家" skill for travel destination substitution.
- Implements unique "体验元素"分析：理解梦想目的地本质体验，再基于预算/时长/签证等约束智能推荐国内外高性价比平替。
- 支持用户偏好记忆与持续成长，提供个性化推荐与方案累积。
- 集成 FlyAI CLI 工具链，支持机票/酒店/景点实时智能搜索。
- 具备匹配度打分、体验透明对比、费用及体验一站式输出等核心功能。
- 包含智能失败处理、季节预算约束判断、自主学习与用户交互提问机制。
Metadata
Frequently Asked Questions
What is 平替旅行家?
平替旅行家——想去A太贵？帮你找平替！AI理解你向往某个目的地的"核心体验元素"，然后在更近/更便宜的目的地中找到能提供相似体验的"平替方案"。当用户提到"想去XX但太贵"、"XX平替"、"找平替"、"去不起XX"、"类似XX的地方"、"XX太远了"、"XX签证难"时使用。 It is an AI Agent Skill for Claude Code / OpenClaw, with 100 downloads so far.
How do I install 平替旅行家?
Run "/install flyai-destination-substitute" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is 平替旅行家 free?
Yes, 平替旅行家 is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does 平替旅行家 support?
平替旅行家 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created 平替旅行家?
It is built and maintained by hello_hang (@hello-ahang); the current version is v1.0.1.
More Skills