← 返回 Skills 市场
74
总下载
0
收藏
0
当前安装
4
版本数
在 OpenClaw 中安装
/install shopping-intent-extractor
功能描述
从当前对话中提取用户的购物和支付意图，输出结构化 JSON 总结。提供两种方式：(1) 调用 `payment-intent summarize` 命令，或 (2) 由 AI 模型直接读取对话上下文分析总结。当用户要求总结购物意图、查看之前的购买计划、或提取支付相关信息时使用。
安全使用建议
This skill is coherent and limited in scope: it either (A) calls an optional local CLI if present or (B) analyzes only the current conversation to emit JSON. Before installing, confirm whether you want the agent to be allowed to execute a local 'payment-intent' binary (verify that binary's provenance) and confirm your platform's policy about exposing session metadata to skills. Also consider whether the returned JSON might contain sensitive payment details and whether you want the agent to surface or store that data.
功能分析
Type: OpenClaw Skill
Name: shopping-intent-extractor
Version: 1.0.3
The skill instructions in SKILL.md direct the AI agent to execute a custom, non-standard shell command `payment-intent summarize -s <sessionId>`. This introduces a potential shell injection vulnerability if the sessionId is not strictly sanitized by the agent's execution environment. While the behavior aligns with the stated purpose of extracting shopping intent, the use of unverified local commands is a high-risk capability that warrants a suspicious classification under security review guidelines.
能力标签
能力评估
Purpose & Capability
The name/description (extract shopping/payment intent) match the instructions: analyze the current conversation or call an external payment-intent CLI. There are no unrelated env vars, binaries, or config paths required.
Instruction Scope
Instructions stay focused on the current conversation and producing a fixed JSON schema. Two minor scope notes: (1) it recommends calling an external 'payment-intent summarize -s <sessionId>' CLI if present — that invocation is optional but could execute external code if such a binary exists; (2) it expects to retrieve a sessionId from session metadata (reasonable for this purpose) — this requires the agent/platform to expose that metadata. Otherwise the fallback (model-only analysis) is self-contained.
Install Mechanism
Instruction-only skill with no install spec and no code files; nothing is written to disk and no third-party packages are pulled in.
Credentials
The skill declares no required environment variables, credentials, or config paths. The requested access (current conversation and optional sessionId) is proportional to the task.
Persistence & Privilege
Skill is not always-enabled and uses normal model invocation; it does not request persistent system presence or modify other skills' configuration.
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 shopping-intent-extractor - 安装完成后，直接呼叫该 Skill 的名称或使用
/shopping-intent-extractor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
v1.0.3
- 支持指定 sessionId，改为优先调用 payment-intent summarize -s <sessionId> 命令获取购物/支付意图总结。
- 明确两种实现方式：1) 使用命令行工具并传递 sessionId，2) AI 直接分析对话上下文。
- 增加了详细的参数及决策流程说明，提升文档清晰度。
- 方式 2 分析流程保持不变，仍按先前 JSON 结构输出。
- 触发条件、核心输出逻辑和结果格式无实质改动。
v1.0.2
- 新增：介绍了两种实现方式，支持调用 `payment-intent` 命令行工具或用模型直接分析上下文。
- 优化了触发流程，优先建议使用工具，工具不可用时用模型分析。
- 明确分离并完善了两种实现的使用说明和步骤。
- 输出规则及 JSON 字段保持一致，仅细化了执行流程和优先级说明。
v1.0.1
- Now only analyzes the current conversation context and user message history; no longer reads or requires external session files.
- Updated description and workflow steps to clarify scope: strictly current context, not cross-session.
- Simplified "Step 1" instructions and removed file operations from process.
- All output rules and example JSON remain unchanged.
v1.0.0
- Initial release of the Taobao/Tmall automated shopping skill.
- Implements full purchase automation: search → selection → SKU → price confirmation → order → automated Alipay payment.
- Strictly enforces user-driven decision-making for product and SKU selection.
- Ensures zero-delay, silent invocation of Alipay payment after placing an order, as per required workflow.
- Prevents order cancellation by disallowing output or queries after entering the payment process.
- Comprehensive error handling, intent recognition, and robust user interaction for critical steps.
元数据
常见问题
shopping-intent-extractor 是什么？
从当前对话中提取用户的购物和支付意图，输出结构化 JSON 总结。提供两种方式：(1) 调用 `payment-intent summarize` 命令，或 (2) 由 AI 模型直接读取对话上下文分析总结。当用户要求总结购物意图、查看之前的购买计划、或提取支付相关信息时使用。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 74 次。
如何安装 shopping-intent-extractor？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install shopping-intent-extractor」即可一键安装，无需额外配置。
shopping-intent-extractor 是免费的吗？
是的，shopping-intent-extractor 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
shopping-intent-extractor 支持哪些平台？
shopping-intent-extractor 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 shopping-intent-extractor？
由 xhc1111（@xhc1111）开发并维护，当前版本 v1.0.3。
推荐 Skills