Terminal-only control improves digital agents across web, desktop and CAD

CUAWright lets agents use bash and the file system to inspect environments, create tools and invoke graphical controls when necessary.

Big Tech
Yadong Lu · Theodore Lee · Yifei Li · Lawrence Keunho Jang · Tianci Xue · Yu Su · +2 more

Microsoft Research · National University of Singapore · The Ohio State University · Carnegie Mellon University

Research Digest··2 min read
Lu and colleagues tested whether computer-use agents benefit from replacing fixed, GUI-centered toolsets with a programmable terminal interface.

CUAWright exposes bash commands as the agent’s sole action format.

Why this paper

From Microsoft Research and 3 others

In one line

A terminal-only harness using bash commands outperforms GUI harnesses across web, desktop, and CAD tasks.

What we could check

  • ·No code link found
  • ·No weights link found
  • ·No dataset link found
  • ·No compute details found
  • ·No stated limitations found
  • ✓Reports numbers on named benchmarks (4 benchmarks)

Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.

§

Research Digest

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.