← Back to Skills Marketplace
106
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install rag-knowledge-curator
Description
企业知识库治理引擎，提供智能分块、去重、标签分类、质量评分与版本管理，解决RAG“垃圾进垃圾出”痛点
Usage Guidance
技能本身是自洽的、低风险的文本处理/治理说明。但注意：
- 不要把敏感凭据或机密（API 密钥、私人数据、未脱敏的机密文档）直接放入 raw_documents——技能会处理并生成可储存的输出，若你将结果自动写入向量数据库，可能导致凭据被持久化。
- 如果你打算把输出“直接对接向量数据库”，请在外部管道（ETL/入库脚本）中使用安全的凭据管理、访问控制与人工复核步骤；确保技能不会被配置为直接将数据推送到不受信任的外部端点。
- 建议先在非敏感样本上测试治理策略与评分规则，确认分块/去重/标签逻辑满足业务需求再在生产数据上运行。
Capability Analysis
Type: OpenClaw Skill
Name: rag-knowledge-curator
Version: 1.0.0
The skill is a text-processing utility designed for RAG (Retrieval-Augmented Generation) data preparation, focusing on cleaning, chunking, and quality scoring of documents. Analysis of SKILL.md, schema.json, and examples.json reveals no evidence of malicious intent, data exfiltration, or harmful prompt injection; the instructions are strictly aligned with the stated purpose of knowledge base governance.
Capability Tags
Capability Assessment
Purpose & Capability
技能名与描述（知识库治理、分块、去重、打标、质量评分、版本化）与 SKILL.md、schema.json、examples.json 中的要求和示例保持一致；没有请求与其目的不符的凭据或二进制依赖。
Instruction Scope
SKILL.md 的运行指令仅在提供的 raw_documents 上执行清洗、分块、抽元数据、评分与生成报告；没有指示读取系统文件或未声明的环境变量。但文档中提到“可直接对接向量数据库入库”，实现细节未给出——如果用户把该输出自动化写入外部存储，需要注意外部管道的安全性和权限控制。
Install Mechanism
无安装规范、无代码、仅为说明文档（instruction-only），不会在主机上下载或写入可执行内容，安装风险低。
Credentials
未声明任何环境变量、凭据或配置路径；所需权限与技能目的相称。
Persistence & Privilege
flags 表示 always:false，默认允许模型自主调用但无强制常驻或跨技能配置修改；没有请求持久化平台级权限。
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install rag-knowledge-curator - After installation, invoke the skill by name or use
/rag-knowledge-curator - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of rag-knowledge-curator.
- Provides enterprise knowledge base curation with cleaning, deduplication, tagging, quality scoring, and version management.
- Supports smart chunking strategies and domain-specific processing.
- Outputs data in markdown format with detailed governance reports and integration recommendations for vector database pipelines.
- Compatible with OpenClaw, Dify, Coze, and SkillHub.
- Designed to address RAG “garbage in, garbage out” challenges from the source.
Metadata
Frequently Asked Questions
What is rag-knowledge-curator?
企业知识库治理引擎，提供智能分块、去重、标签分类、质量评分与版本管理，解决RAG“垃圾进垃圾出”痛点. It is an AI Agent Skill for Claude Code / OpenClaw, with 106 downloads so far.
How do I install rag-knowledge-curator?
Run "/install rag-knowledge-curator" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is rag-knowledge-curator free?
Yes, rag-knowledge-curator is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does rag-knowledge-curator support?
rag-knowledge-curator is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created rag-knowledge-curator?
It is built and maintained by boboy-j (@boboy-j); the current version is v1.0.0.
More Skills