HarnessCompass: Guiding Automatic Harness Evolution toward Generalizable and Effective Agent Harnesses

2026-08-03

AI Summary

Existing methods for automatic harness evolution often suffer from overfitting, reliance on a single trajectory signal, and interference among jointly optimized components, which collectively limit their generalization capability. This work proposes HarnessCompass, a novel framework that enhances generalization through task-agnostic global constraints, incorporates first-person active feedback from the agent, and decouples harness components to enable independent optimization, thereby minimizing mutual interference. Evaluated on SWE-bench Verified, HarnessCompass improves the Pass@1 score of GPT-5.4 from 54% to 66% within just five iterations, significantly outperforming AHE. Moreover, it demonstrates superior generalization on both unseen tasks and models.

Abstract

Harness design plays a critical role in agent performance by shaping how large language models (LLMs) perceive, reason over, and act within executable environments. Recent work has proposed automatic harness evolution, which iteratively improves the harness from agent--environment interactions. However, existing methods often overfit to the evolution tasks, rely exclusively on trajectory-derived signals, and optimize harness components jointly, causing interference across components. We propose HarnessCompass, a novel automatic harness evolution framework built around constrained evolution, proactive feedback, and component-wise optimization. HarnessCompass first enforces global constraints on evolution, restricting modifications to task-agnostic harness changes that generalize beyond the evolution tasks. It then augments trajectory-derived evidence with proactive first-person feedback from the agent about harness usage, yielding richer signals for evolution. Finally, it decouples the optimization of different harness components before consolidating them into a unified harness, reducing cross-component interference while preserving component synergy. On SWE-bench Verified with GPT-5.4, HarnessCompass improves Pass@1 from 54\% to 66\% in only 5 evolution iterations, outperforming AHE in both effectiveness and evolution efficiency. In addition, the evolved harness transfers effectively to held-out tasks and other models, demonstrating substantially stronger generalization than prior automatic harness evolution methods.

Problem

• automatic harness evolution
• overfitting
• trajectory-derived signals
• component interference
• generalization

Innovation

• automatic harness evolution
• constrained evolution
• proactive feedback
• component-wise optimization
• generalization