JIT-Agent：透過即時框架演進擴展框架智能

摘要

智能體能力並非僅由模型本身決定。涵蓋記憶管理、規劃策略、動作協議與工具/技能編排的智能體框架，其影響力可能遠超底層基礎模型的貢獻。然而，框架設計至今仍依賴人工、針對特定任務，且本質上無法規模化。我們提出 JIT-Agent——一個框架智能模型，經訓練可為任意現成的智能體 LLM 即時合成任務自適應的智能體框架。我們將智能體框架形式化為可組合、機器可生成的產物，受固定四模組協議規範，並訓練 JIT-Agent 為指定任務定制框架、修復框架以確保穩定可靠的執行，以及透過從不斷擴充的既有框架配置存檔中蒸餾效能訊號來自我演化。配備 JIT-Agent 作為框架輔助工具，DeepSeek-V4-Flash 在 DeepSearchQA（+9.1）與 OdysseyBench（+4.3）上超越 GPT-5.6，而本已強大的 GLM-5.2 更獲得高達 +20.2 分的提升。在受控評估中，JIT-Agent 生成的框架在效能上可與 OpenCode 和 Claude Code 等成熟的智能體執行環境相抗衡，並持續提升 DeepSeek V4、Mimo-V2.5 與 Qwen3.6 等多規模模型系列的表現。據我們所知，JIT-Agent 是首個專為即時框架生成而設計的模型，確立了框架智能作為一種可訓練、可遷移且可累積的智能體能力維度，與模型規模擴展相互正交。

English

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.