skills.rest
saddam-eng-tech
saddam-eng-techCOMMUNITY
Explore all skills by this author →
GitHub

rag-gap-auditor

Execute structured test queries to audit RAG document corpora for gaps.

Updated Jan 17, 2026
Search Tags:#llm#rag#knowledge base#retrieval#auditing#gap analysis
One-click install
npx skills add https://github.com/saddam-eng-tech/ai-agent-skills --skill rag-gap-auditor
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-gap-auditor
Source: https://github.com/saddam-eng-tech/ai-agent-skills/tree/main/rag-gap-auditor
Command: npx skills add https://github.com/saddam-eng-tech/ai-agent-skills --skill rag-gap-auditor

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill identifies and helps remediate knowledge gaps in Retrieval-Augmented Generation (RAG) systems, ensuring more accurate and relevant information retrieval.

Core Features & Use Cases

  • Systematic Gap Probing: Runs structured test queries across various categories to uncover retrieval failures.
  • Document Corpus Auditing: Checks documents for staleness, poor chunking, and missing coverage.
  • Prioritized Remediation Plan: Delivers actionable steps ranked by impact to improve RAG performance.
  • Use Case: When your RAG system is failing to find relevant documents or providing incorrect answers, this skill will diagnose the root cause and provide a clear plan to fix it.

Quick Start

Use the rag-gap-auditor skill to audit my RAG system with these failing queries.

Frequently Asked Questions about rag-gap-auditor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
Why does my RAG system fail to retrieve relevant documents for specific queries?▼

RAG retrieval failures often stem from knowledge gaps, document staleness, or poor chunking. This skill diagnoses these issues by executing structured test queries and auditing the document corpus to identify root causes.

How do I audit my knowledge base for missing coverage and stale information?▼

You can audit a knowledge base by probing it with structured test queries and analyzing the document corpus. This skill identifies staleness, poor chunking, and missing coverage, then provides a prioritized remediation plan.

What is the best way to find root causes of incorrect answers in retrieval-augmented generation?▼

Finding root causes of incorrect RAG answers requires systematic failure analysis. This skill classifies retrieval failures by executing test queries across categories and auditing embeddings and document chunking strategies.

Do I need to understand embedding models to diagnose RAG retrieval failures?▼

Yes, diagnosing retrieval failures requires an understanding of RAG components, embedding models, and retrieval strategies. This knowledge is necessary to interpret the structured test queries and prioritize the remediation plan.

How do I fix poor chunking and stale documents in my LLM knowledge base?▼

To fix poor chunking and stale documents, run a corpus audit to classify failure root causes. This skill provides a prioritized remediation plan with actionable steps ranked by impact to improve RAG performance.

Can I systematically test RAG knowledge gaps across different query categories?▼

Yes, you can systematically test RAG knowledge gaps by executing structured test queries across various categories. This process uncovers retrieval failures and checks the document corpus for missing coverage.

Related Skills

Explore More Data & Analytics Skills→
nist0nist0

rag-patterns

Optimize RAG pipeline chunking, embedding, and retrieval strategies.

Community
Advanced
agentsopeagentsope
287

agentsop-idempotent-ingestion

Make RAG ingestion idempotent with a persisted content-hash ledger.

Community
Advanced
lagarcesslagarcess
1

rag-architect

Design and optimize retrieval-augmented generation pipelines for document indexing and vector search.

Community
Advanced
simplysmartaisimplysmartai

llm-evaluation

Evaluate LLM applications with automated metrics, human feedback, and LLM-as-judge patterns.

Community
Advanced