住宿顾问

围绕已知地标通过 POI 锚定和半径筛选查找并对比附近酒店。适用于用户明确景点、商圈、交通枢纽或会展中心，并希望按距离与质量排序获得可预订住宿。

Security Scan

The skill's requirements and runtime instructions are consistent with a hotel-finding assistant that drives an external CLI (flyai-cli); nothing in the package suggests it is trying to do unrelated or hidden actions — but it depends on installing a third-party npm CLI and network access, so verify that CLI before installing.

Purpose & Capability

Name/description (find and compare hotels around a POI) match the instructions: the SKILL.md explicitly calls out using flyai-cli commands (search-poi, search-hotels) and describes expected inputs/outputs. The declared need for Node and a flyai-cli is consistent with the stated purpose.

Instruction Scope

Runtime instructions are scoped to locating POIs and hotels, filtering by radius/budget/stars, and formatting results. They do not instruct reading unrelated files, scanning system state, or exfiltrating arbitrary secrets. The only external interactions are CLI calls and network access to the fly.ai/Fliggy endpoints described.

Install Mechanism

This is an instruction-only skill (no install spec), but SKILL.md instructs the user to run `npm i -g @fly-ai/flyai-cli`. Installing a third‑party global npm package runs remote code and can be risky if the package or its maintainer are untrusted. The skill does not provide a pinned package version or a verified source (GitHub repo link), so users should inspect the package before installing.

Credentials

No required environment variables are declared. The only optional credential mentioned is an optional FLYAI_API_KEY for improved results, which is appropriate for a tool that calls an external API. There are no unrelated credential requests or config-path access in the instructions.

Persistence & Privilege

The skill does not request always:true, does not include install-time scripts or files, and is instruction-only. It does not attempt to modify other skills or system settings. Autonomous invocation is enabled (the platform default) but not combined with other high-risk factors here.

Assessment

This skill appears to be what it says: an adapter that uses a third‑party CLI (flyai-cli) to find and format nearby hotels. Before installing or running it: 1) Inspect the npm package (@fly-ai/flyai-cli) — check its npm page, source repo, maintainer, version history, and recent downloads; avoid installing global packages you don't trust. 2) Be aware that queries (city, landmark, dates) will be sent over the network to fly.ai / Fliggy endpoints; do not include sensitive personal data you don't want transmitted. 3) If you want stronger assurance, run the CLI in a sandbox/container or review its source code first. 4) The optional FLYAI_API_KEY grants the CLI access to your FlyAI account — only provide keys you trust and understand the API's permissions.

何时使用

• 用户已明确目标地标（景点、商圈、车站、机场或会展中心）。
• 用户对当地地理不熟悉，希望以地标为锚点快速圈定住宿范围。
• 目标是输出可比较、可预订的酒店候选，而不是泛化的旅行建议。

输入/输出约定

• 最小输入：city、landmark keyword、check-in date、check-out date。
• 增强输入：radius、budget、hotel stars、sorting preference。
• 最小输出：推荐结论、对比表、酒店图片行、预订链接行。
• 链接约束：酒店预订链接必须使用 detailUrl 字段。

快速开始

1. 安装 CLI：npm i -g @fly-ai/flyai-cli
2. 验证环境：执行 flyai fliggy-fast-search --query "三亚有什么可玩" 并确认输出 JSON。
3. 查看命令：执行 flyai --help。
4. 查看参数说明：阅读 references/ 中文档，确认必填/选填参数与字段映射。

配置说明

该工具可在无 API Key 情况下试用。若需要增强效果，可配置可选密钥：

flyai config set FLYAI_API_KEY "your-key"

核心流程（最小可执行路径）

1. 确认地标锚点（POI）：
   • flyai search-poi --city-name "<city>" --keyword "<landmark>"
2. 拉取地标周边酒店：
   • flyai search-hotels --dest-name "<city>" --poi-name "<landmark>" --check-in-date <yyyy-mm-dd> --check-out-date <yyyy-mm-dd> --sort <distance_asc|rate_desc> --max-price <budget>
3. 执行半径筛选：
   • 若有经纬度：按 distanceKm <= radiusKm 计算并过滤
   • 若无经纬度：保留 distance_asc 排序，并标注“近似半径”

输出规则（强约束）

• 使用 Markdown 标题、要点和对比表。
• 若存在图片字段，先输出图片行：![]({mainPic|picUrl})
• 然后输出预订链接行：
  • search-hotels -> [Click to book]({detailUrl})
  • 其他 -> [Click to book]({jumpUrl})

详细参考

• 场景化方案与参数速查：references/playbooks.md
• 问询清单、参数模板与输出模板：references/templates.md
• 失败兜底与重试策略：references/fallbacks.md
• 执行日志模板：references/runbook.md