Agentic Router: An Execution-Grounded Continual Learning Approach With Memory
2026-08-10 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors present a system that helps computers manage network commands more safely and effectively. Their approach produces several possible commands, predicts what might happen when each is run, and then picks the best one based on expected benefits and risks. They improve this by learning from past command runs to suggest better commands and make smarter choices over time. Tests showed their method leads to more commands that work correctly on the first try.
Large Language ModelsCommand-Line InterfaceNetwork OperationsSONiCExecution FeedbackConsequence PredictionAction SelectionLoRA (Low-Rank Adaptation)SSHMulti-turn Sessions
Authors
Yuxuan Chen, Rongpeng Li, Zhifeng Zhao, Yuntao Liu, Xing Xu, Honggang Zhang
Abstract
Large language model (LLM) agents provide a promising interface for command-line-based network operations, but a plausible command may still fail or introduce operational risk after execution. Existing approaches mainly focus on command generation or final configuration correctness, and do not use execution-grounded experience to jointly improve candidate coverage and action selection. We propose an execution-grounded dual-path consequence-aware agent for CLI-based SONiC operations, which generates multiple complete actions, predicts their execution consequences, and selects the final action through utility- and risk-aware reranking. The proposal-side path abstracts reusable operational lessons into retrievable guidance to improve feasible-action coverage without modifying the proposal LLM, while the selection-side path adapts the consequence predictor through session-level LoRA updates using real SSH feedback to improve conditional selection quality. Experiments over multi-turn SONiC operation sessions with different Qwen3 proposal models show that the framework improves feasible-action coverage and top-1 execution success, and that the two adaptation paths provide complementary gains over interaction.