← 返回 Skills 市场

AI Enterprise Knowledge Base

Name: AI Enterprise Knowledge Base
Author: yang1002378395-cmyk

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

cross-platform ⚠ suspicious

229

总下载

当前安装

版本数

在 OpenClaw 中安装

/install ai-enterprise-knowledge-base

功能描述

快速搭建企业内部问答系统，支持多格式文档上传、智能语义检索、权限管理和多平台接入。

使用说明 (SKILL.md)

AI Enterprise Knowledge Base

5分钟搭建企业内部问答系统。

功能特性

✅ 文档上传：支持 PDF/Word/Markdown/Excel ✅ 智能检索：向量搜索 + 语义理解 ✅ 权限管理：部门/角色/用户三级权限 ✅ 多渠道接入：飞书/企业微信/钉钉/Slack ✅ 知识图谱：自动提取实体关系 ✅ 数据分析：问答统计/知识缺口分析

快速开始

1. 安装 OpenClaw

git clone https://github.com/openclaw/openclaw.git
cd openclaw && npm install

2. 配置 Skill

name: ai-enterprise-knowledge-base
config:
  # 文档存储路径
  storage: ./knowledge-base
  
  # 向量数据库（可选）
  vectorDb: milvus  # or pinecone/weaviate
  
  # LLM 配置
  llm:
    provider: deepseek
    model: deepseek-chat
    
  # 权限配置
  auth:
    enabled: true
    provider: ldap  # or oauth/saml

3. 启动服务

openclaw skills install ai-enterprise-knowledge-base
openclaw start

使用示例

上传文档

curl -X POST http://localhost:3000/api/documents \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -F "[email protected]" \
  -F "department=hr"

问答

curl -X POST http://localhost:3000/api/ask \
  -H "Authorization: Bearer YOUR_TOKEN" \
  -d "question=公司的年假政策是什么？"

飞书集成

# config/channels.yaml
- platform: feishu
  app_id: cli_xxx
  app_secret: xxx
  events:
    - message.receive

定价

套餐	价格	功能
基础版	¥99	文档上传 + 智能检索
专业版	¥299	权限管理 + 多渠道接入
企业版	¥999	知识图谱 + 数据分析 + 定制开发

企业案例

某科技公司（500人）

知识库文档：5000+
月均问答：50000+
新员工培训时间：缩短 60%
内部工单：减少 40%

某电商公司（200人）

客服知识库：自动回答 80% 常见问题
客服响应时间：从 5 分钟降至 10 秒
客户满意度：提升 25%

技术支持

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

安装配置服务：¥99 起，30 分钟搞定！

安全使用建议

This skill mostly does what it says (enterprise QA server) but take precautions before installing: 1) Inspect the OpenClaw GitHub repository referenced in the SKILL.md before running git clone/npm install so you know what code will execute. 2) Note package.json in the bundle references index.js but no code is included — ask the author or prefer installing from the upstream project rather than trusting the bundle. 3) Prepare any integration secrets (LLM keys, vector DB creds, Feishu app_secret, LDAP/OAuth settings) and avoid pasting them into public repos; confirm where and how they are stored. 4) If possible, install and run the service in an isolated environment (VM/container) and review network activity during startup. 5) Ask the publisher for a homepage/source and a reproducible install path if you need higher assurance.

功能分析

Type: OpenClaw Skill Name: ai-enterprise-knowledge-base Version: 1.0.0 The skill bundle contains standard metadata and documentation for an AI-driven enterprise knowledge base. The SKILL.md file provides typical installation instructions, configuration examples, and API usage guides without any evidence of prompt injection, malicious commands, or data exfiltration. While the package.json refers to an index.js that is not present in the provided files, the existing content is purely informational and aligns with the stated purpose of the skill.

能力评估

ℹ Purpose & Capability

The name/description (enterprise KB with uploads, vector search, auth, channel integrations) matches the SKILL.md instructions and examples. However, the bundle contains only SKILL.md and package.json (no index.js or other code), so the package metadata suggests runnable code that isn't included — this mismatch is unexplained and worth verifying.

✓ Instruction Scope

Runtime instructions stay within the stated purpose: they show how to install OpenClaw, configure storage/vector DB/LLM/auth, start the service, upload docs, and configure channel integrations. The instructions do require you to provide tokens/secrets (Authorization bearer, feishu app_secret) but do not instruct reading unrelated system files or exfiltrating data.

⚠ Install Mechanism

There is no formal install spec in the skill metadata, but SKILL.md tells users to git clone https://github.com/openclaw/openclaw.git && npm install — this pulls and executes third‑party code from the network. That is expected for a full application but increases risk: you should inspect the upstream repo before running install. The skill bundle itself contains no installable code despite package.json declaring a main file.

ℹ Credentials

The skill metadata declares no required env vars, which is consistent with being instruction-only. The instructions, however, expect you to supply credentials and secrets in configs (LLM provider credentials, vector DB credentials, Feishu app_secret, LDAP/OAuth/SAML configs). Those are proportional for integrations but the skill does not explicitly document which secrets are stored where or whether they are transmitted to third parties — verify how credentials are managed before use.

✓ Persistence & Privilege

Skill flags are default: not always-included and user-invocable. The skill doesn't request elevated platform privileges in the metadata and does not claim to modify other skills or system-wide agent settings.

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-enterprise-knowledge-base
安装完成后，直接呼叫该 Skill 的名称或使用 /ai-enterprise-knowledge-base 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: 5分钟搭建企业内部问答系统

元数据

Slug ai-enterprise-knowledge-base

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