VibeBench/VibeSearchBench
VibeSearchBench introduces a multi-turn search-agent benchmark with verifiable knowledge-graph evaluation and early GitHub traction.
Excerpt
🔍 The hardest search benchmark in the wild — vague, multi-turn, proactive. 200 long-horizon tasks with persona-driven progressive disclosure, scored by verifiable schema-free knowledge-graph evaluation. No vibes, just triplet F1. — ★ 573 · Python · topics: agentic-ai, benchmark, llm, proactive-agent, search, search-agent
Read at source: https://github.com/VibeBench/VibeSearchBench