5 projects
cua-auto-harness
Conformance harness for computer-use agent stacks: verifies that intended interactions actually happen
autoresearch-kernel
The autoresearch kernel: experiments as folders (seed + rules + objective), agents and runners as swappable pillars.
programming-with-pixels
Programming with Pixels (PwP) - A framework for computer-use software engineering agents
automix-llm
Library for Mixing Multiple Models with differential costs and performance.
AdaptiveConsistency
Library for running AdapativeConsistency based Inference on large language models.
