Token导航 LogoToken导航TokenDH.com
效率需要联网clawhub未标认证来源可访问clear审计提醒

retail-agent-setup零售 Agent 设置

Agent Skill

retail-agent-setup 用于辅助前端页面、组件、样式和交互逻辑开发,适合在 OpenClaw 中需要维护前端项目、生成组件或检查界面实现时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

3,305

周安装

135

GitHub Stars

公开资料未说明

下载量

1,069
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:retail-agent-setup(零售 Agent 设置)
来源仓库:https://github.com/fangwei-frank/retail-agent-setup
安装命令:
openclaw skills install retail-agent-setup
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install retail-agent-setup

简介

零售数字员工代理的标准化部署向导,分步指导完成 12 项核心配置任务。

  • 帮助企业快速搭建可操作的 AI 商店助理,降低技术门槛。
  • 提供参数校验、接口对接与基础功能测试的一体化流程指引。
  • 执行前请准备必要的商户 ID、API 权限与数据存储路径信息。
  • 注意:生产环境部署前建议先在沙箱环境验证各环节兼容性。

SKILL.md

name
retail-agent-setup
description
>
metadata
openclaw
emoji
🏪

Retail Agent Setup — Onboarding Wizard

Overview

This skill transforms a blank OpenClaw agent into a fully configured retail digital employee tailored to a specific store or chain. Each step produces a concrete artifact that persists in the agent's memory, making the setup cumulative and resumable.

Setup takes 20–40 minutes end-to-end. Each step can be paused and resumed. Run retail agent setup or 数字员工配置 to start or continue.


Execution Protocol

  • Run steps in order — each step depends on outputs from the previous
  • Pause after each step — show the artifact, ask "Confirm and continue?" before proceeding
  • Resumable — if a step was previously completed, show its saved output and ask whether to redo or skip
  • Save state — write each step's output to agent memory before moving to the next
  • Zero-config entry — if the user just says "set up my retail agent," start at Step 1

The 12 Steps

Step 01 — System Inventory

"What retail systems are you currently using?"

Identify the store's existing tech stack across 5 categories: POS, ERP/WMS, CRM/membership, e-commerce platforms, and supply chain tools.

Map each system to its API availability (real-time / batch / none). Reference: step-01-systems.md Artifact: System inventory card + API availability matrix


Step 02 — Data Infrastructure Assessment

"Where does your data live, and what format is it in?"

Evaluate data across 6 dimensions: products, inventory, sales, staff, customers, and policy docs. Score completeness and freshness. Prioritize what to connect first. Reference: step-02-data-infra.md Artifact: Data map + connection priority list


Step 03 — Data Import & Auto-Structuring

"Send me your data — I'll organize it into a format the agent can use."

Accept uploads (Excel/CSV/PDF/Word/image), API connections, or pasted text. Auto-parse into structured knowledge base entries. Flag gaps and prompt to fill them. Script: scripts/parse_products.py — Excel/CSV → structured JSON Script: scripts/parse_policy.py — PDF/Word → rule tree Script: scripts/score_knowledge.py — completeness scoring Reference: step-03-data-import.md Artifact: Structured knowledge base + completeness score (0–100)


Step 04 — Role Selection

"What role should this digital employee play?"

Choose from 6 preset roles or define a custom role. Each role activates a specific skill bundle and response style. One agent = one primary role (multi-role is advanced config). Reference: step-04-role-select.md Artifact: Role definition file + activated skill bundle list


Step 05 — Skills Configuration

"Which capabilities should this agent have?"

Review recommended skills for the chosen role. Toggle on/off. Configure each enabled skill (thresholds, data sources, escalation rules). Reference: step-05-skills-config.md Artifact: skills-config.json — active skills with their parameters


Step 06 — Knowledge Base Validation

"Let me test what your agent knows."

Auto-generate 10 test questions covering products, inventory, policies, and recommendations. Run them against the knowledge base. Flag failures. Guide the user to fill gaps. Script: scripts/gen_test_cases.py — generate test questions by vertical Script: scripts/score_knowledge.py — run and score responses Reference: step-06-knowledge.md Artifact: Knowledge base score + gap report


Step 07 — Digital Employee Persona

"Give your digital employee a name and personality."

Configure: name, personality type, tone, reply style, customer address form, brand keywords. Generate 3 sample dialogues for preview. Confirm before saving. Reference: step-07-persona.md Artifact: persona-config.json + 3 preview dialogues


Step 08 — Channel Integration

"How will staff and customers reach this agent?"

Select and configure delivery channels: WeCom (企业微信), WeChat MP/Mini Program, Lark (飞书), Web kiosk UI, WhatsApp, or SMS/IVR. Each channel has a dedicated setup guide with step-by-step auth instructions. Reference: step-08-channels.md Artifact: Channel connection status + test message confirmation


Step 09 — Permissions & Escalation

"What can the agent decide alone, and what needs a human?"

Define 4-level permission matrix: L0 auto-handle, L1 suggest+confirm, L2 submit for approval, L3 force escalate to human. Set escalation targets and on-call schedules. Reference: step-09-permissions.md Artifact: permissions-matrix.json + escalation routing config


Step 10 — Pre-Launch Testing

"Let's run real-scenario tests before going live."

Run a full scenario test suite based on the store's vertical and configured skills. Score readiness 0–100. Must reach 80+ to proceed to launch. Script: scripts/gen_test_cases.py Reference: step-10-test.md Artifact: Test report + launch-readiness score


Step 11 — Launch & Handoff

"You're ready. Let's go live."

Activate the agent on all configured channels. Generate staff onboarding card (one-pager). Send welcome message. Schedule first check-in reminder (7 days out). Reference: step-11-handoff.md Artifact: Staff guide PDF + activation confirmation


Step 12 — Continuous Improvement

"Going live is the beginning, not the end."

Set up weekly unanswered-question digests and monthly usage reports. Configure knowledge-gap alerts. Schedule quarterly persona review. Reference: step-12-iterate.md Artifact: Cron jobs for digest + alert thresholds set


State Management

Track onboarding progress in agent memory under key retail_setup_state:

{
  "version": "1.0",
  "started_at": "<ISO timestamp>",
  "completed_steps": [1, 2, 3],
  "current_step": 4,
  "artifacts": {
    "systems": { ... },
    "data_map": { ... },
    "knowledge_base": { ... },
    "role": "...",
    "skills_config": { ... },
    "persona": { ... },
    "channels": [ ... ],
    "permissions": { ... }
  }
}

On any new message, check this state first. If setup is incomplete, offer to resume.


Supported Retail Verticals

Apparel · Footwear · Beauty & Skincare · Consumer Electronics · Home & Furniture · Maternal & Infant · Convenience Store · Supermarket · Specialty Food · Jewelry · Sporting Goods · Books & Stationery · Pet Supplies · Pharmacy · Toy & Hobby

For verticals not listed, use "General Retail" defaults and customize in Step 4.

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

95.24%
按下载量换算1,018

安全审计

VirusTotal

通过

ClawScan

可疑

Static analysis

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills