skills.rest
Gzmomo001
Gzmomo001COMMUNITY
Explore all skills by this author →
WebsiteGitHub

rag-retrieval-debugger

Diagnose and fix hybrid retrieval pipeline issues in antifraud RAG systems.

3|1|Updated Apr 10, 2026
Search Tags:#logging#debug#retrieval#BM25#vector-search#RRF#antifraud
One-click install
npx skills add https://github.com/Gzmomo001/antichet_RAG --skill rag-retrieval-debugger
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rag-retrieval-debugger
Source: https://github.com/Gzmomo001/antichet_RAG/tree/main/.claude/skills/rag-retrieval-debugger
Command: npx skills add https://github.com/Gzmomo001/antichet_RAG --skill rag-retrieval-debugger

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Diagnose and fix issues in the hybrid retrieval pipeline (BM25 + embeddings + RRF) for antifraud RAG systems.

Core Features & Use Cases

  • Diagnose and reproduce retrieval issues across BM25, embedding, and RRF fusion stages.
  • Verify service health by testing the embedding service, BM25 queries, and vector search, then apply targeted fixes.
  • Use case: when analyze endpoints return unexpected results or scores, follow the debugging workflow to isolate and resolve the problem.

Quick Start

Start a debugging session for a failing retrieval query and follow the workflow to diagnose and fix the hybrid search.

Frequently Asked Questions about rag-retrieval-debugger

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

FAQPage Schema
How do I debug unexpected RAG retrieval results in an antifraud system?▼

To debug unexpected hybrid retrieval results, follow a structured workflow to diagnose and fix issues across BM25, embedding generation, and RRF fusion stages. This verifies service health and isolates failures within the antifraud RAG pipeline.

What is the best way to troubleshoot incorrect BM25 matching in a RAG pipeline?▼

Troubleshooting incorrect BM25 matching involves testing the BM25 query service directly and logging outputs to identify scoring anomalies. The debugging workflow verifies BM25 matching independently before checking vector search and RRF fusion.

Why does my vector search return irrelevant embeddings for antifraud analysis?▼

Vector search returns irrelevant embeddings when the embedding generation service fails or produces incorrect vectors. Diagnose this by testing the embedding service health and logging vector search outputs to verify embedding quality before RRF fusion.

How do I fix RRF fusion scoring issues in hybrid retrieval?▼

Fix RRF fusion scoring issues by testing the individual BM25 and vector search stages first to ensure correct baseline inputs. The debugging workflow isolates whether RRF fusion or its upstream retrieval components cause the unexpected scores.

Can I use this debugging workflow for development and investigation workflows?▼

Yes, the structured debugging workflow applies across both investigation and development workflows. It helps diagnose and fix retrieval issues when analyze endpoints return unexpected results or incorrect scores in the antifraud RAG system.

Related Skills

Explore More Software Engineering Skills→
ivanshamaevivanshamaev
14

rag-data-pipeline

Build RAG ingestion and query pipelines with hybrid retrieval and reranking.

Community
Advanced
JSai23JSai23
1

vault-sem-search-a

Enable semantic search of local vault notes using qmd with BM25, vector search, and LLM re-ranking.

Community
Intermediate
latestaiagentslatestaiagents
5

hybrid-retrieval

Combine dense vector embeddings with sparse BM25 retrieval for RAG systems.

Official
Advanced
ionmidoriionmidori

implementing-hybrid-search

Combine vector and keyword search results using Reciprocal Rank Fusion.

Community
Intermediate