← 返回 Skills 市场

AI Customer Service Automation

Name: AI Customer Service Automation
Author: yang1002378395-cmyk

作者 yang1002378395-cmyk · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ⚠ suspicious

355

总下载

当前安装

版本数

在 OpenClaw 中安装

/install ai-customer-service-automation

功能描述

24/7智能客服系统，自动回复常见问题，情感分析升级工单，多渠道支持，数据看板和知识库同步。

使用说明 (SKILL.md)

AI Customer Service Automation

7x24小时智能客服系统。

核心能力

✅ 智能问答：自动回复 80% 常见问题 ✅ 情感分析：识别愤怒/失望/满意，自动升级 ✅ 工单系统：复杂问题自动创建工单 ✅ 多渠道支持：飞书/微信/钉钉/Telegram/邮件 ✅ 数据看板：响应时间/解决率/满意度 ✅ 知识库同步：自动学习历史对话

快速开始

1. 安装

openclaw skills install ai-customer-service-automation

2. 配置

name: ai-customer-service-automation
config:
  # 知识库
  knowledgeBase:
    files:
      - ./docs/faq.md
      - ./docs/policy.pdf
      
  # 情感阈值
  sentiment:
    alertThreshold: 0.3  # 愤怒 \x3C 0.3 自动升级
    
  # 工单系统
  ticket:
    provider: jira  # or feishu/dingtalk
    autoCreate: true
    
  # 多渠道
  channels:
    - feishu
    - wechat
    - telegram

3. 启动

openclaw start

ROI 计算

传统客服 vs AI 客服

指标	传统客服	AI 客服	提升
响应时间	5-10 分钟	10 秒	30-60x
人力成本	¥5000/月/人	¥500/月	10x
工作时间	8 小时	24 小时	3x
并发能力	1 对 1	无限	∞
满意度	70%	85%	+15%

定价

套餐	价格	功能
基础版	¥199	智能问答 + 1 渠道
专业版	¥499	情感分析 + 工单 + 3 渠道
企业版	¥1499	全功能 + 定制开发

客户案例

某电商平台

日均咨询：5000+
AI 自动处理：4000+
人工介入：20%
客服成本：降低 60%

某 SaaS 公司

用户数：10000+
AI 回答准确率：92%
客户满意度：4.5/5
客服团队：从 10 人降至 3 人

技术支持

📧 Email: [email protected]
💬 Telegram: @openclaw_service
📱 微信: openclaw-cn

安装配置服务：¥199 起，1 小时搞定！

安全使用建议

This bundle looks like a product brochure + config template rather than a complete implementation. Before installing or enabling it: 1) Ask the author for the missing source code (index.js/connector code) and for an explicit list of required credentials and environment variable names. 2) Do not provide broad/shared API keys — use least-privileged, service-specific tokens if you proceed. 3) Verify how personal or customer data is handled (where conversation logs and KB files are stored, retention, and external endpoints). 4) If connectors are provided later, review their code for hardcoded endpoints or data exfiltration, and prefer self-hosted or vetted integration components. 5) If you only want documentation, keep it disconnected from runtime; do not paste production credentials into the config until you audit the integration code.

功能分析

Type: OpenClaw Skill Name: ai-customer-service-automation Version: 1.0.0 The provided files consist solely of metadata and documentation for an AI customer service automation skill. There is no executable code, and the instructions in SKILL.md are purely descriptive of the service's features and configuration without any malicious prompt injection or harmful commands.

能力评估

ℹ Purpose & Capability

The SKILL.md describes multi-channel support (Feishu/WeChat/DingTalk/Telegram/email), Jira ticketing, sentiment analysis, KB sync and gives a config example — that matches the declared purpose. However package.json names index.js as main but no code files are present, and the skill declares no required environment variables or credentials even though integrating with those services would require API keys/tokens. This mismatch (claimed capabilities without code or declared connectors/credentials) is inconsistent and could mean the skill is just documentation or is incomplete.

ℹ Instruction Scope

Runtime instructions are high-level: install the skill, provide a YAML config (including local KB file paths), and start. The instructions only reference local KB files and standard openclaw commands — they do not tell the agent to read unrelated system files or to exfiltrate data. However they are vague about how to connect to external channels and how credentials will be provided/used, leaving broad discretion and potential privacy/behavior gaps.

✓ Install Mechanism

No install spec and no code files beyond package.json — instruction-only skill. That is low install risk (nothing is downloaded or written by an installer). Note: package.json references an index.js that is not present; that suggests the package is incomplete or missing implementation.

⚠ Credentials

The skill claims integrations that normally require multiple credentials (Jira API token, Feishu/DingTalk/WeChat bot tokens, Telegram bot token, email SMTP credentials, etc.), yet requires.env is empty and SKILL.md does not list or name any environment variables. This omission is disproportionate and ambiguous: either the skill expects the platform to provide credentials implicitly, or the author omitted critical security details. That ambiguity increases risk because users might be prompted later to supply sensitive tokens without clear guidance.

✓ Persistence & Privilege

The skill does not request always:true and is user-invocable with normal autonomous invocation allowed. There is no indication it modifies other skills or agent-wide configs. No persistence/privilege red flags in the provided manifest.

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 ai-customer-service-automation
安装完成后，直接呼叫该 Skill 的名称或使用 /ai-customer-service-automation 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

v1.0.0

Initial release: 7x24h智能客服系统，支持多渠道、情感分析、工单系统

元数据

Slug ai-customer-service-automation

版本 1.0.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题