rag-engineer
Build Reliable RAG Systems
RAG projects often fail because retrieval quality is weak or unmeasured. This skill guides chunking, embeddings, vector search, hybrid retrieval, reranking, and evaluation choices.
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Review the Skillstore skill "rag-engineer" from https://skillstore.io/skills/sickn33-rag-engineer.md and its manifest at https://skillstore.io/api/skills/sickn33-rag-engineer/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "rag-engineer". Design RAG for a support knowledge base with product manuals and tickets.
Expected outcome:
- Recommended semantic chunks aligned to headings and troubleshooting steps.
- Metadata filters for product, version, language, and document type.
- Hybrid retrieval with reranking before sending context to the model.
Using "rag-engineer". Our RAG answers are plausible but miss exact policy details.
Expected outcome:
- Likely retrieval issue caused by weak metadata filtering or oversized chunks.
- Add policy section identifiers, rerank top candidates, and evaluate retrieval separately.
- Use relevance thresholds to avoid filling the prompt with weak matches.
Using "rag-engineer". Evaluate whether retrieval quality improved after changing embeddings.
Expected outcome:
- Build a query set with expected source passages and edge cases.
- Measure recall, precision, reranker impact, and failure categories.
- Compare results by content type before updating production embeddings.
Security Audit
SafeAll static findings are false positives. The file contains markdown guidance for RAG design, and the flagged backticks are code fences or inline formatting. No prompt injection, exfiltration intent, or executable command behavior was found.
Risk Factors
⚙️ External commands (7)
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/sickn33-rag-engineer/audits/4?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/sickn33-rag-engineer?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/sickn33-rag-engineer?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/sickn33-rag-engineer/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/sickn33-rag-engineer.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
sickn33. (2026). rag-engineer security audit report (audit version 4) [Author version unspecified]. Skillstore. https://skillstore.io/skills/sickn33-rag-engineer/audits/4BibTeX citation
@techreport{sickn33-sickn33-rag-engineer-2026,
author = {sickn33},
title = {rag-engineer security audit report (audit version 4)},
institution = {Skillstore},
year = {2026},
number = {4},
url = {https://skillstore.io/skills/sickn33-rag-engineer/audits/4},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "rag-engineer security audit report (audit version 4)"
version: "unspecified"
type: report
authors:
- name: "sickn33"
date-released: "2026-07-07"
url: "https://skillstore.io/skills/sickn33-rag-engineer/audits/4"
identifiers:
- type: other
value: "skillstore:sickn33-rag-engineer:audit:4"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: HighWhat You Can Build
Plan a Knowledge Base RAG System
Create a retrieval architecture for internal documents, including chunking, metadata, hybrid search, and quality checks.
Improve Search Relevance
Diagnose weak retrieval results and tune chunking, embeddings, reranking, filters, and relevance thresholds.
Evaluate Retrieval Quality
Separate retrieval evaluation from generation quality so teams can measure recall, precision, and context usefulness.
Try These Prompts
Help me design a RAG system for these documents: [document types]. Recommend chunking, metadata, embedding, and retrieval choices.
Review my chunking plan for [content type]. Identify risks with boundaries, overlap, metadata, and context preservation.
My RAG system uses vector search and keyword search. Suggest a hybrid retrieval and reranking plan for [query patterns].
Create an evaluation plan for my RAG retriever. Include test data, metrics, failure analysis, and refresh checks for changing documents.
Best Practices
- Preserve document structure and metadata during chunking.
- Use hybrid retrieval when exact terms and semantic meaning both matter.
- Measure retrieval quality separately before judging generated answers.
Avoid
- Using fixed-size chunks that split sentences or sections.
- Indexing every document without filtering low-value content.
- Sending first-stage retrieval results directly to the model without reranking.