ContextIQ: A Reproducible Benchmark for Token-Efficient, Verifiable AI Code Context
收藏官方服务：
资源简介：
ContextIQ is a local code-graph server that gives AI coding agents token-efficient, verifiable context. This deposition archives its reproducible benchmark: retrieval quality (Recall@5=0.989 over 96 queries across 4 corpora), cross-language test-discovery (F1=0.9474), and a hallucination-guard measurement, plus a content-hashed dataset manifest for exact reproduction.
提供机构：
Zenodo创建时间：
2026-07-30
