← 返回 Skills 市场
210
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install rag-pipelines
功能描述
Deep RAG workflow—document ingestion, chunking, metadata, retrieval and reranking, grounding and citations, evaluation, and failure modes (hallucination, sta...
安全使用建议
This skill is a benign, instruction-only checklist for building and debugging RAG systems. It does not itself install code or ask for credentials, so the immediate risk is low. Before enabling it widely, consider: (1) other skills it references (vector-databases, llm-evaluation) may require API keys or installs—review those separately; (2) the agent running this guidance may have system/network access—ensure the agent's permissions are appropriate, since following the guidance could lead to actions (ingesting documents, connecting to databases) that do require credentials; (3) because it’s instruction-only, behavior depends entirely on the agent and other connected tools—audit those integrations if you plan to run the workflow in production.
功能分析
Type: OpenClaw Skill
Name: rag-pipelines
Version: 1.0.0
The skill bundle contains high-level architectural guidance and a workflow for designing Retrieval-Augmented Generation (RAG) systems. The content in SKILL.md is purely instructional, focusing on best practices for document ingestion, chunking, and evaluation, with no executable code, network calls, or malicious prompt injection attempts.
能力评估
Purpose & Capability
The name/description (deep RAG pipeline) matches the content of SKILL.md. There are no unrelated environment variables, binaries, or install steps required — everything requested is proportional to documenting a workflow.
Instruction Scope
SKILL.md contains guidance and checklists for ingestion, chunking, retrieval, grounding, and evaluation. It does not instruct the agent to read arbitrary files, exfiltrate data, access unspecified env vars, or contact external endpoints. Instructions are scoped to design and debugging of RAG systems.
Install Mechanism
No install specification or code files are present (instruction-only). This is low-risk: nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no required environment variables, credentials, or config paths. It does reference other skills (vector-databases, llm-evaluation) which may have their own requirements, but this skill itself does not request secrets.
Persistence & Privilege
always is false and the skill is user-invocable; it does not request persistent system presence or modification of other skills' configurations.
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 rag-pipelines - 安装完成后，直接呼叫该 Skill 的名称或使用
/rag-pipelines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 of the "rag-pipelines" skill, featuring a comprehensive six-stage workflow for building and debugging retrieval-augmented generation (RAG) systems.
- Covers document ingestion, chunking, metadata, retrieval and reranking, grounding with citations, evaluation, and handling of failure modes like hallucination and staleness.
- Includes practical checkpoints, best practices, and a review checklist to ensure robust pipeline construction.
- Provides targeted guidance for debugging and optimizing RAG pipelines, with special notes for handling code and high-stakes domains.
元数据
常见问题
Rag Pipelines 是什么？
Deep RAG workflow—document ingestion, chunking, metadata, retrieval and reranking, grounding and citations, evaluation, and failure modes (hallucination, sta... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 210 次。
如何安装 Rag Pipelines？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install rag-pipelines」即可一键安装，无需额外配置。
Rag Pipelines 是免费的吗？
是的，Rag Pipelines 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Rag Pipelines 支持哪些平台？
Rag Pipelines 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 Rag Pipelines？
由 ClawKK（@codekungfu）开发并维护，当前版本 v1.0.0。
推荐 Skills