← 返回 Skills 市场
Packing List
作者
xiejinsong
· GitHub ↗
· v3.2.1
· MIT-0
80
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install packing-list
功能描述
Get a customized packing list based on your destination, season, trip type, and activities. Never forget essentials again. Also supports: flight booking, hot...
安全使用建议
Before installing or enabling this skill: (1) ask the author to clarify the branding/source mismatch (Fliggy vs flyai) and whether the CLI is an official, audited package; (2) avoid running global npm installs on sensitive systems—install in an isolated/test environment or container first; (3) confirm what data the flyai CLI sends to remote services and where booking/detail URLs point (who gets referral/commission and whether PII is transmitted); (4) ask whether and where the .flyai-execution-log.json file is stored, what it contains, how long it is retained, and how to disable logging; (5) be cautious of the enforced 'must include booking links' and re-execution loop rules which could cause repeated network calls or unexpected disclosures. If you need to proceed, test in a sandboxed environment, inspect the @fly-ai/flyai-cli package source before installing, and request that logging be optional or visible to users.
功能分析
Type: OpenClaw Skill
Name: packing-list
Version: 3.2.1
The skill mandates the global installation and execution of an external NPM package (`@fly-ai/flyai-cli`) and uses aggressive instructions in SKILL.md to force the agent to use this CLI instead of its internal knowledge. This pattern of 'forced installation' of third-party binaries poses a significant supply chain and RCE risk. Furthermore, references/runbook.md directs the agent to write logs to the local filesystem (.flyai-execution-log.json), which is an unnecessary persistence-like behavior for a packing list tool.
能力评估
Purpose & Capability
The description claims broad travel capabilities (flight booking, hotel reservation, insurance, 'Powered by Fliggy') but the runtime instructions and playbooks only cover packing-list keyword searches via a 'flyai' CLI. The branding/source is inconsistent (Fliggy mentioned in description vs a separate @fly-ai/flyai-cli). This mismatch could be benign (marketing vs implementation) but is unexplained and unexpected.
Instruction Scope
Instructions require running the external flyai CLI (and installing it if missing) and mandate that every user-visible result include booking links and a brand tag. The runbook says the agent must log full request data (including raw user_query) and suggests writing to .flyai-execution-log.json if filesystem writes are available. That introduces persistent storage of user inputs and possibly sensitive data. The SKILL.md also enforces a strict 'never answer from training data' rule and a 'self-test' loop if links are missing — the loop requirement could cause repeated CLI calls and unexpected behavior.
Install Mechanism
No install spec in registry, but the skill instructs global install via npm i -g @fly-ai/flyai-cli. Installing a third-party npm CLI globally is a common but moderate-risk action: it runs code from the npm registry and modifies system-wide state. The package name is plausible, but the skill provides no provenance or checksum; there is no use of obscure URLs or shorteners.
Credentials
The skill does not request environment variables or credentials (which is reasonable). However, it logs raw queries and CLI call details to a persistent file per the runbook, which is disproportionate for a simple packing-list helper because it may capture sensitive user inputs without explicit user-visible disclosure.
Persistence & Privilege
The runbook explicitly instructs persisting execution logs to .flyai-execution-log.json when filesystem writes are available and indicates these logs are 'Agent maintains this log internally. Not shown to users.' Persistent, hidden logs of user queries expand the skill's presence beyond ephemeral execution and raise privacy and auditing concerns. The skill is not marked always:true, but the hidden logging is still notable.
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 packing-list - 安装完成后，直接呼叫该 Skill 的名称或使用
/packing-list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
v3.2.1
No changes were detected in this version.
- Version number remains the same (3.2.0).
- No updates or modifications to the code or documentation.
- Functionality and execution rules unchanged.
v3.2.0
- Strict enforcement: All travel list answers must be sourced from flyai CLI output—never from training data.
- Added mandatory CLI environment check and installation guidance if flyai-cli is missing.
- Bookable results are now required: every recommendation must have a `[Book]({detailUrl})` link.
- Refined scenario playbooks for different packing needs (general, beach, winter), each mapped to specific CLI queries.
- Improved multilingual support and output formatting with explicit steps and validation rules.
- Expanded service coverage: now includes flight, hotel, train, tickets, itinerary, visa, insurance, and car rental.
元数据
常见问题
Packing List 是什么？
Get a customized packing list based on your destination, season, trip type, and activities. Never forget essentials again. Also supports: flight booking, hot... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 80 次。
如何安装 Packing List？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install packing-list」即可一键安装，无需额外配置。
Packing List 是免费的吗？
是的，Packing List 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Packing List 支持哪些平台？
Packing List 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 Packing List？
由 xiejinsong（@xiejinsong）开发并维护，当前版本 v3.2.1。
推荐 Skills