Search papers, labs, and topics across Lattice.
This paper introduces JIT-Agent, a novel harness intelligence model that autonomously synthesizes task-adaptive agent harnesses for various off-the-shelf agentic LLMs, addressing the limitations of manual and task-specific harness design. By formalizing the agent harness as a composable artifact governed by a four-module protocol, JIT-Agent can customize, repair, and self-evolve harnesses based on performance signals from previous configurations. The results show that JIT-Agent significantly enhances the performance of models like DeepSeek-V4-Flash and GLM-5.2, outperforming state-of-the-art benchmarks and establishing harness intelligence as a critical, scalable dimension of agent capability.
JIT-Agent redefines agent performance by enabling on-the-fly harness evolution, leading to substantial improvements over existing models.
Agent capability is not determined by the model alone. The agent harness, encompassing memory management, planning strategy, action protocol, and tool/skill orchestration, can dominate the contribution of the underlying foundation model. Yet harness design remains manual, task-specific, and fundamentally unscalable. We present JIT-Agent, a harness intelligence model trained to synthesize task-adaptive agent harnesses on the fly for arbitrary off-the-shelf agentic LLMs. We formalize the agent harness as a composable, machine-generatable artifact governed by a fixed four-module protocol, and train JIT-Agent to customize harnesses for a given task at hand, repair harnesses for stable and reliable execution, and self-evolve by distilling performance signals from an expanding archive of prior harness configurations. Equipped with JIT-Agent as a harness helper, DeepSeek-V4-Flash surpasses GPT-5.6 on DeepSearchQA (+9.1) and OdysseyBench (+4.3), while the already strong GLM-5.2 gains up to +20.2 points. Across controlled evaluations, JIT-Agent-generated harnesses are performance-competitive with mature agent runtimes such as OpenCode and Claude Code and consistently improve multi-scale model families of DeepSeek V4, Mimo-V2.5, and Qwen3.6. To our knowledge, JIT-Agent is the first model purpose-built for just-in-time harness generation, establishing harness intelligence as a trainable, transferable, and compounding dimension of agent capability orthogonal to model scaling.
