GitHub· suyoumo·· 2026-08-26AI 评分26
ClawProBench：面向 OpenClaw 运行时的 LLM 智能体评测基准
suyoumo/ClawProBench — ClawProBench is a live-first benchmark harness for evaluating LLM agents in the OpenClaw runtime with deterministic grading and repeated-trial reliability.
AI 导读
ClawProBench 是一个面向 OpenClaw 运行时的 LLM 智能体评测基准工具，采用 live-first 方式，提供确定性评分与重复试验可靠性。该基准用于在 OpenClaw 运行时中评估 LLM 智能体表现。
来源：GitHub · github.com