Changelog

v0.1.0 (2026-06-13)

• Initial evaluation suite release
• 100-item golden eval dataset from 5 open-license RPG sources
• Retrieval metrics: Recall@k, MRR, NDCG@k
• LLM-as-judge generation scoring
• Regression testing with baseline comparison
• Latency benchmarking harness
• Branded SVG badge generation
• CLI via python -m raghelm eval
• CI integration with GitHub Actions