Agentic rag routing

Skill Goodnight77/rag-skills/skills/rag-agents/agentic-rag-routing

Route each query to the right retrieval strategy, data source, or tool with an LLM-driven router.From its SKILL.md

Install

npx -y skills add Goodnight77/rag-skills --skill agentic-rag-routing

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

7 stars7 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.3 KB, as published. Nobody here has run it

Overview

Agentic routing puts a decision step in front of retrieval: the agent classifies the query and chooses whether to retrieve, which source or index to hit, and which tool to call. This turns a one-size-fits-all pipeline into one that adapts per query.

Problem Statement

Different queries need different handling. A definitional question, a multi-hop question, and a live-data question all fail under a single fixed retrieval path. Without routing, the pipeline over-retrieves for simple queries and under-serves complex ones.

Key Concepts

Retrieval decision: Whether the query needs retrieval at all.
Source routing: Selecting the right index, collection, or tool for the query.
Query classification: Mapping a query to a handling strategy.
Fallback: A safe default path when the router is unsure.

Implementation Guide

Step 1: Define the Route Space

Enumerate the concrete routes available, such as no-retrieval, vector search, keyword search, web search, or a structured-data tool.

Why: A router can only choose from routes you have defined; an open-ended router is unpredictable and hard to evaluate.

Step 2: Classify the Query

Use an LLM or a lightweight classifier to map each query to a route, returning a structured decision rather than free text.

Why: Structured routing decisions are testable and can be logged, replayed, and evaluated.

Step 3: Add Fallback and Guardrails

Define a default route for low-confidence decisions and cap how many tools or hops a single query may trigger.

Why: Guardrails prevent runaway tool loops and keep latency and cost bounded.

For patterns, see the LangGraph agentic RAG guide and the LlamaIndex router query engine.

When to Use This Skill

Queries span multiple sources, formats, or freshness needs
Some queries need no retrieval and others need several hops
You want per-query control before adding full multi-agent orchestration

When NOT to Use This Skill

All queries are answered well by one retrieval path
The extra classification call breaks the latency budget
There is only one data source and no tool choice to make

Related Skills

Adaptive Retrieval - Strategy selection by query type
Self-RAG - Retrieval decisions at generation time
CRAG - Corrective RAG - Fallback search on weak retrieval
Multi-Agent RAG Orchestration - When routing is not enough

Metrics & Success Criteria

Route Accuracy: Queries reach the correct handling path
No-Retrieve Savings: Simple queries skip unnecessary retrieval
Bounded Cost: Tool and hop counts stay within limits
Fallback Rate: Low-confidence routing stays rare and safe