Xiangyi Li
Source-listed: Founder · BenchFlow (Benchmarkthing Inc.)
Record snapshot: Sep 3, 2026. Coverage and affiliations may be incomplete or historical.
Selected work
5- SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse TasksSaved credit: First author · 2026Explore people connected to this work →
- SkillsBench benchmark, tasks, and evaluation repositorySaved credit: Contributor · 2026Explore people connected to this work →
- BenchFlowSaved credit: CreatorExplore people connected to this work →
- ClawsBench: Evaluating Capability and Safety of LLM Productivity Agents in Simulated WorkspacesSaved credit: Contributor · 2026Explore people connected to this work →
- ClawsBenchSaved credit: Contributor · 2026Explore people connected to this work →