← 返回 Skills 市场
105
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install eo-ability-rag
功能描述
RAG知识共享能力,跨项目共享专家经验和最佳实践,基于语义搜索快速检索
使用说明 (SKILL.md)
eo-ability-rag
RAG 知识共享能力 - 跨项目共享知识,避免重复造轮子
一句话介绍
基于RAG架构的知识共享能力,跨项目共享专家经验和最佳实践,语义搜索快速检索。
核心功能
- 知识索引: 自动索引专家经验和最佳实践
- 语义搜索: 自然语言查询相关知识
- 经验共享: 跨项目共享成功经验
- 持续学习: 从使用中自动更新知识库
使用方法
索引知识
# 自动索引(由 EO 自动触发)
索引: "Architect 专家在博客系统项目中的架构设计经验"
索引: "最佳实践:电商平台的缓存策略"
# 手动索引
/索引 "安全审计的 PCI-DSS 检查清单"
搜索知识
# 搜索相关知识
搜索: "博客系统的架构设计模式"
搜索: "学术论文的方法论设计"
搜索: "电商营销的内容策略"
# 基于专家搜索
搜索: "小红书策略师的种草内容经验"
共享知识
# 标记为可共享
共享: "本次项目中学到的架构设计经验"
来源: "博客系统项目"
专家: "Architect"
与EO插件的协同
- 被所有 eo-workflow-* 调用
- 使用 LanceDB(插件版)或简单索引(独立版)
- 被案例3(电商营销运营)使用
独立运行模式(有EO vs 无EO)
| 模式 | 能力 |
|---|---|
| 有EO插件 | LanceDB向量索引、语义搜索、跨会话知识持久化 |
| 无插件(基础) | 简单关键词索引、基础全文搜索 |
示例
// 用户需要设计博客系统架构
const knowledge = await eo_ability_rag({
action: 'search',
query: '博客系统架构设计最佳实践'
})
// → 返回相关的架构模式、设计经验
// 使用知识辅助设计
const arch = await eo_ability_architect({
task: '设计博客系统',
useKnowledge: knowledge.results
})
Interface
Input
interface RAGInput {
action: 'index' | 'search' | 'share' | 'learn' | 'stats'
content?: string // 要索引的内容
query?: string // 搜索查询
expert?: string // 专家领域
project?: string // 项目名称
}
Output
interface RAGOutput {
results: KnowledgeEntry[]
relevance: number[]
sources: string[]
totalIndexed: number
}
🦞⚙️ 钢铁龙虾军团
安全使用建议
This skill appears coherent and low-risk as provided (instruction-only, no creds). Before installing: confirm what '自动索引' will read (documents, conversations, repos), where the vector DB (LanceDB) would be hosted, and what credentials/storage that plugin needs. Do not allow automatic indexing of sensitive secrets or private files until you know the ingestion sources and retention/ACLs. If you require privacy, ask for explicit details about the storage backend, access controls, and how to delete indexed entries.
功能分析
Type: OpenClaw Skill
Name: eo-ability-rag
Version: 1.0.0
The skill bundle defines a Retrieval-Augmented Generation (RAG) capability for indexing and searching project-related knowledge. The content in SKILL.md and _meta.json consists entirely of documentation, interface definitions, and usage examples without any executable code, external network requests, or suspicious instructions.
能力评估
Purpose & Capability
Name, description, and the SKILL.md all describe a RAG-based knowledge-sharing/search capability. There are no unrelated required binaries, env vars, or config paths. References to LanceDB and an 'EO' plugin are consistent with a vector-index/search implementation.
Instruction Scope
Instructions remain within the stated purpose (index/search/share/learn). However the document is high-level and vague about what '自动索引' (automatic indexing) consumes, how data is ingested, and what is persisted. It implies autonomous/background indexing and cross-session persistence, so it's important to confirm what data sources and triggers are used before indexing sensitive information.
Install Mechanism
No install spec and no code files — instruction-only skill. This minimizes disk/writing risk because nothing in the skill will be installed or executed by itself.
Credentials
The skill declares no environment variables or credentials, which is proportionate. It does mention optional use of LanceDB (plugin) or '独立版' simple index; the SKILL.md does not specify whether using LanceDB requires external credentials, endpoints, or storage. Verify what the EO/LanceDB integration requires before enabling it in an environment with sensitive data.
Persistence & Privilege
The skill does not request always:true and is user-invocable. It claims '持续学习' and cross-session persistence; combined with autonomous invocation this implies the agent or plugin may store and update a knowledge base over time. Confirm retention, access controls, and whether shared entries can be exported or visible to other projects.
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 eo-ability-rag - 安装完成后,直接呼叫该 Skill 的名称或使用
/eo-ability-rag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
eo-ability-rag v1.0.0
- Initial release of a RAG-based knowledge sharing capability.
- Enables cross-project sharing of expert experiences and best practices.
- Features semantic search for rapid retrieval of relevant knowledge.
- Supports both automated and manual knowledge indexing.
- Integrates with EO plugins and supports standalone operation with varying search functions.
元数据
常见问题
Eo Ability Rag 是什么?
RAG知识共享能力,跨项目共享专家经验和最佳实践,基于语义搜索快速检索. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 105 次。
如何安装 Eo Ability Rag?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install eo-ability-rag」即可一键安装,无需额外配置。
Eo Ability Rag 是免费的吗?
是的,Eo Ability Rag 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Eo Ability Rag 支持哪些平台?
Eo Ability Rag 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 Eo Ability Rag?
由 467718584(@467718584)开发并维护,当前版本 v1.0.0。
推荐 Skills