← Back to Skills Marketplace
85
Downloads
1
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install must-play-recommendations
Description
为用户推荐城市的必玩景点。当用户询问某个城市有什么好玩的、景点推荐、必去景点、旅游攻略时触发。核心能力是基于 FlyAI CLI 获取城市的5A级景区和热门好评景点，提供详细推荐信息。
Usage Guidance
Before installing or enabling this skill: (1) Understand it relies on the FlyAI CLI — ensure flyai is installed from a trusted source (e.g., official package registry) or the skill will fail. (2) The FlyAI CLI can use an API key (FLYAI_API_KEY) or stored config; confirm whether you have such credentials configured and whether you want this skill to make networked queries using them. If you don't want it to access your FlyAI account, do not set the API key in the environment or CLI config. (3) Ask the skill author to declare required binaries and any credential usage in the registry metadata (so there are no surprises). (4) If you care about isolation, run the skill in a sandboxed environment or review FlyAI CLI call outputs (they include external booking links) before acting on returned URLs. (5) If anything about undeclared dependencies or network calls makes you uncomfortable, treat this as moderate-risk and request the maintainer add an install spec and explicit credential handling documentation.
Capability Analysis
Type: OpenClaw Skill
Name: must-play-recommendations
Version: 1.0.1
The skill provides travel recommendations by executing shell commands via the `flyai` CLI tool, as seen in `SKILL.md`. It constructs these commands using user-provided input (e.g., `flyai search-poi --city-name "{城市名称}"`), which introduces a potential shell injection vulnerability if the agent fails to sanitize the input. While the instructions include a step to validate the city name and the behavior aligns with the stated travel-agent purpose, the reliance on shell execution for data retrieval is a high-risk capability that warrants a suspicious classification under the provided criteria.
Capability Assessment
Purpose & Capability
The skill's name/description (city attraction recommendations) matches the instructions: it uses FlyAI CLI commands (search-poi, ai-search) to gather POI and semantic results. However, the SKILL.md expects the flyai CLI to be present and package.json lists @fly-ai/flyai-cli as a dependency, yet the registry metadata declares no required binaries and there is no install spec — an inconsistency between claimed runtime needs and declared requirements.
Instruction Scope
Runtime instructions remain within the stated purpose: extract city name, call flyai search-poi and ai-search, merge and de-duplicate results, and output markdown. The instructions do not ask the agent to read unrelated system files or transmit arbitrary local data. They will include external booking links returned by FlyAI (expected for travel recommendations).
Install Mechanism
This is instruction-only with no install spec (lowest install risk). Nevertheless package.json lists the FlyAI CLI dependency but no automated install is provided; the skill assumes the environment already has flyai installed or the operator will install it. That mismatch should be resolved or documented to avoid silent failures or surprise manual installs.
Credentials
The skill declares no required environment variables, but the FlyAI CLI (per flyai-readme.md) supports/uses an optional FLYAI_API_KEY and configuration. If a user or environment has FlyAI credentials configured, the skill will indirectly use them via the CLI. Because credentials are not declared in metadata and the SKILL.md doesn't explicitly state how secrets are used, there's a risk of undeclared credential access/exfiltration via CLI network calls. The skill itself does not request unrelated credentials, but the indirect dependency on a networked CLI raises proportionality concerns.
Persistence & Privilege
The skill is not always-enabled and does not request persistent/privileged platform presence. It's instruction-only and does not modify other skills or system-wide settings according to available files.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install must-play-recommendations - After installation, invoke the skill by name or use
/must-play-recommendations - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
- 新增 FlyAI CLI 详细文档链接（flyai-readme.md）作为相关资源。
- 功能和逻辑未变，仅补充文档说明与资源引用，完善使用指导。
v1.0.0
- Initial release of the must-play-recommendations skill.
- Provides detailed, high-quality travel recommendations for any specified Chinese city, focusing on 5A and highly rated attractions.
- Integrates FlyAI CLI’s search-poi and ai-search commands to gather real-time data on 5A/4A scenic spots and top-reviewed destinations.
- Outputs results in rich Markdown format including images, descriptions, booking links, and travel tips.
- Strict rules exclude low-quality, unpopular, or solely “cost-effective” recommendations.
- Includes robust error handling and extensible design for future data sources (e.g., Xiaohongshu, Dianping).
Metadata
Frequently Asked Questions
What is must-play-recommendations?
为用户推荐城市的必玩景点。当用户询问某个城市有什么好玩的、景点推荐、必去景点、旅游攻略时触发。核心能力是基于 FlyAI CLI 获取城市的5A级景区和热门好评景点，提供详细推荐信息。 It is an AI Agent Skill for Claude Code / OpenClaw, with 85 downloads so far.
How do I install must-play-recommendations?
Run "/install must-play-recommendations" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is must-play-recommendations free?
Yes, must-play-recommendations is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does must-play-recommendations support?
must-play-recommendations is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created must-play-recommendations?
It is built and maintained by SeaSon (@ysxiiun); the current version is v1.0.1.
More Skills