Skip to content
Arxo Arxo

Examples and Report Walkthrough

Examples and Report Walkthrough

Section titled “Examples and Report Walkthrough”

This page explains how to interpret rag_architecture current output.

Sample Artifacts

Section titled “Sample Artifacts”
  • crates/arxo-engine/src/metrics/ai_observability/rag_architecture/samples/chat-langchain-report.json
  • crates/arxo-engine/src/metrics/ai_observability/rag_architecture/samples/chat-langchain-example-report.md

Start with composites

Section titled “Start with composites”
  • rag_architecture.retrieval_readiness
  • rag_architecture.safety_readiness
  • rag_architecture.evaluation_readiness
  • rag_architecture.overall_health

Triage by category

Section titled “Triage by category”
  • Retrieval: retrieval_latency_risk, context_budget_gap, reranking_gap, retrieval_thresholding_gap, query_planning_gap
  • Safety: retrieval_scope_filter_gap, prompt_injection_guard_gap, retrieved_content_sanitization_gap, retrieval_trust_boundary_gap
  • Evaluation/Grounding: eval_harness_gap, eval_metric_coverage_gap, grounding_citation_gap, citation_precision_gap, citation_coverage_gap

Prioritization pattern

Section titled “Prioritization pattern”
  1. Fix high-severity safety gaps first.
  2. Fix retrieval quality/performance gaps that affect serving paths.
  3. Tighten evaluation and citation quality until evaluation_readiness_score stabilizes.
Section titled “Read Next”