skills.rest
MerkyorLynn
MerkyorLynnCOMMUNITY
Explore all skills by this author →
GitHub

rag-assistant

Retrieve relevant passages from an indexed local knowledge base and synthesize source-grounded answers.

37|4|Updated Mar 31, 2026
Search Tags:#information retrieval#rag#knowledge base#conflict resolution#citation#document qa#web fallback
One-click install
npx skills add https://github.com/MerkyorLynn/Lynn --skill rag-assistant
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-assistant
Source: https://github.com/MerkyorLynn/Lynn/tree/main/plugins/rag-core/skills/rag-assistant
Command: npx skills add https://github.com/MerkyorLynn/Lynn --skill rag-assistant

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you get accurate answers grounded in your own previously indexed documents, instead of relying on generic web or memory-based responses.

Core Features & Use Cases

  • Knowledge-base retrieval with relevance checking: It queries your local knowledge index and detects when the results are not actually related, then falls back to web search.
  • Source-grounded, conflict-aware answers: It requires citing the exact source file for each fact and summarizes differences when multiple documents disagree.
  • Clear fit boundaries: It supports document-based questions and comparisons, while explicitly steering users to knowledge_index when the documents are not indexed.

Quick Start

Ask rag-assistant to answer your question based on your indexed documents and cite the supporting source files.

Frequently Asked Questions about rag-assistant

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

FAQPage Schema
How do I answer questions from my own indexed documents?▼

To answer questions from indexed documents, retrieve relevant passages by querying a local knowledge base, then synthesize a response with per-fact citations to the exact source files.

What happens when document retrieval doesn't find relevant information?▼

When document retrieval relevance is insufficient, the system falls back to web search to find the needed information, while still noting any conflicting viewpoints found across sources.

Can I compare conflicting information across multiple private files?▼

Yes, you can compare conflicting information across private files by extracting entities and querying the knowledge base, which summarizes differences when multiple documents disagree.

Do I need to index my files before asking document Q&A questions?▼

Yes, you must index files using a knowledge indexing tool first, because this system only answers questions based on private files that were already indexed in the local knowledge base.

How are knowledge-base claims separated from general knowledge in responses?▼

Knowledge-base claims are separated from general knowledge by enforcing per-fact citations to source files, ensuring every piece of retrieved information is explicitly grounded in your indexed documents.

Related Skills

Explore More Education & Research Skills→
muzhicaomingwangmuzhicaomingwang
3

zhimeng-agent

Answer Obsidian vault questions and generate daily Feishu reports via RAG.

Community
Advanced
alo-expalo-exp
6

performance-report

Generate structured marketing performance reports from cross-channel data.

Official
Advanced
artificial-intelligence-firstartificial-intelligence-first

SKILL.md Guide (Agent Skills)

Package domain knowledge into modular units for AI agents.

Official
Advanced
alexanderopalexanderop
15

adding-notes

Convert resource URLs into structured Markdown notes with frontmatter and wiki-links.

Community
Basic