# amazon-science/RAGChecker

RAGChecker: A Fine-grained Framework For Diagnosing RAG

Repository: https://github.com/amazon-science/RAGChecker
Canonical: https://ross.abutalabs.com/products/ragchecker
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2024-12-13T13:24:41+00:00

## Health v2 (maintenance only)
Score: 25/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 40, longevity 57
- inputs: {"age_days": 800, "days_push": 628, "days_rel": 707, "gap_med": 0, "n_releases_24m": 6}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1110, forks 92 (observed 2026-08-28T04:03:37.486894+00:00)

## What it is
RAGChecker is an automatic evaluation framework for diagnosing Retrieval-Augmented Generation (RAG) systems. It provides holistic and diagnostic metrics for retrieval and generation components, using claim-level entailment for fine-grained analysis, plus a benchmark dataset.

## Use cases
- evaluate my rag pipeline quality
- diagnose whether retrieval or generation is failing in my rag system
- benchmark rag systems across domains
- measure faithfulness and relevance of rag answers
- compare rag checker metrics with human judgments
- run claim-level entailment evaluation on rag outputs

## When to choose
- you need fine-grained, component-level diagnosis of a RAG pipeline
- you want automated metrics correlated with human judgments
- you need a benchmark dataset for RAG evaluation research

## When to avoid
- you need to build or serve a RAG system rather than evaluate one
- you need a lightweight evaluation without LLM-based entailment costs
- you need non-English evaluation support

## Facets
- artifact type: framework
- maturity: active
- function: rag, benchmarking, testing, nlp, llm-inference, cli
- domain: large-language-models, machine-learning, developer-tools
- platform: python, cli, cross-platform
- tags: rag-evaluation, diagnostic-metrics, claim-level-entailment, evaluation-framework, retrieval-evaluation, generator-evaluation, retrieval-augmented-generation, natural-language-processing

## Member repositories
- amazon-science/RAGChecker (main) score 25

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:37.486894+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:43:26.315280+00:00, confidence not recorded.
  - readme: https://github.com/amazon-science/RAGChecker (fetched 2026-08-28T04:03:37.486894+00:00, sha d2a5e03773b4)
- Data as of 2026-08-30T08:39:29.467469+00:00.