HarnessBridge: Learnable Bidirectional Controller for LLM Agent Harness
For developers of LLM-based agents, this work addresses the scalability bottleneck of manually engineered harnesses by providing a learnable alternative that generalizes across models.
HarnessBridge introduces a learnable bidirectional controller for LLM agent harnesses, achieving comparable or better performance than specialized harnesses on Terminal-Bench 2.0 and SWE-bench Verified while reducing token usage and trajectory length.
Large language models are increasingly deployed as agents for long-horizon tasks, yet their performance is shaped not only by model capability and environment design, but also by the harness that mediates agent--environment interaction. Existing harnesses are largely manually engineered, making them difficult to scale as trajectories grow longer and interactions become more complex. In this work, we ask whether harness can be generated by a learnable plug-in module that can be trained in an end-to-end fashion. We introduce HarnessBridge, a lightweight learnable harness controller that parameterizes the agent--environment interface as a bidirectional projection. HarnessBridge learns two bidirectional projections: observation projection, which distills raw trajectories into compact, decision-relevant states, and action projection, which converts proposed actions into executable transitions or trajectory-grounded rejections. We train HarnessBridge on a harness supervision dataset via unified instruction tuning. On Terminal-Bench~2.0 and SWE-bench Verified, HarnessBridge matches or surpasses strong specialized harnesses while substantially reducing token usage and trajectory length, and generalizes from smaller generators to larger commercial models.