Abstract
Self-improvement in AI agents has emerged as a key research frontier: systems that modify their own prompts, workflows, and decision rules based on accumulated operational experience. The state-of-the-art Self-Harness framework achieves a 14-21% improvement on Terminal-Bench-2.0 by mining failure clusters and patching the agent harness. However, Self-Harness optimizes only one dimension -- the prompt harness -- leaving behavioral principles and workflow topology unchanged.
We propose APEX (Adaptive Principle EXtraction), a three-layer co-evolution framework that simultaneously evolves:
- Layer 1: the harness via failure-mode patching.
- Layer 2: behavioral principles via success-trace distillation.
- Layer 3: the agent workflow topology via structural fitness-based selection.
We implement APEX on Joe, a production-grade super AI Agent built on NVIDIA Nemotron, designed as an Edge AI Agent Factory for the NVIDIA Agent Challenge 2026, managing a 15-node compute fleet using 114 real task traces collected over 18 days. APEX achieves an APEX Health Score of 0.570 (+90% vs. baseline 0.300) in a single evolutionary run, distilling 6 novel reusable principles and selecting a research-first workflow topology scoring 0.900 (+20%). Our results demonstrate that multi-dimensional co-evolution substantially outperforms single-axis harness optimization, at a cost of only 4 LLM calls (~270 s) on a local qwen2.5-coder:32b instance.
Blogger's Review: The APEX framework significantly enhances the performance of AI agents through multi-dimensional self-evolution, offering new design insights for future intelligent systems. The method strikes a commendable balance between efficiency and effectiveness, showcasing the immense potential of self-improvement in practical applications.