rag-eval-harness
LLM-judge evaluation report · 3 items · judge: FakeJudge (deterministic heuristic, zero-cost demo) · generated 2026-08-30T19:22:41.121Z
Faithfulness66.7%
Answer Relevance96.3%
Context Precision50.0%
Context Recall66.7%
| ID | Question | Faithfulness | Relevance | Precision | Recall |
|---|---|---|---|---|---|
| q1 | What port does the default HTTP server listen on? | 100.0% | 100.0% | 100.0% | 100.0% |
| q2 | How does the retriever rank documents? | 100.0% | 100.0% | 50.0% | 100.0% |
| q3 | What database does the project use for vector storage? | 0.0% | 88.9% | 0.0% | 0.0% |