A smarter agent harness for TerminalBench, built on top of Terminus 2
Simple fixes, significant gains.
uv run harbor run \
--dataset terminal-bench-sample@2.0 \
--n-tasks 1 \
--agent-import-path "terminus_kira.terminus_kira:TerminusKIRA" \
--model anthropic/claude-opus-4-6 \
--env docker \
-n 1For more details, visit our blog post.
If you found Terminus-KIRA useful, please cite us as:
@misc{terminuskira2026,
title={Terminus-KIRA: Boosting Frontier Model Performance on Terminal-Bench with Minimal Harness },
author={{KRAFTON AI} and {Ludo Robotics}},
year={2026},
url={https://github.com/krafton-ai/kira},
}KRAFTON AI & Ludo Robotics