Research

Environments are the new data.

Models in 2026 don’t get better from more static prompts. They get better from running through realistic environments and being judged on the whole workflow. Our research artifacts ship benchmarks, environments, and the runtime that runs both.

Jan 2026Benchmark
SkillsBench
Skills enable agent deployment across domains. The first benchmark for whether procedural skills — instructions, scripts, references an agent loads on demand — make agents better at real work. 86 tasks, 11 domains.
Xiangyi Li, Han-chung Lee, +39 contributors
Apr 2026Environment
ClawsBench
Wire-compatible mock workplaces for agents. Five mock workplaces — Gmail, Calendar, Drive, Docs, Slack — wire-compatible with the upstream `gws` and Slack APIs. Production agents and skills run unchanged against a safety-evaluable replica.
Xiangyi Li, Kyoung Whan Choe, Yimin Liu, +12 contributors
Late 2024 — ActiveRuntime
BenchFlow
Models are the cars. We build the track.. The agent simulation runtime. One Scene-based lifecycle for single-agent, multi-agent, and multi-round evals. Sandboxed, hardened against reward hacking, full trajectory capture.
Xiangyi Li, Yimin Liu, Han-chung Lee

Environments are the new data.

SkillsBench

ClawsBench

BenchFlow