Adapting LLM agents to domain-specific tool calling remains notably brittle under evolving interfaces. Prompt and schema engineering is easy to deploy but often fragile under distribution shift and strict parsers, while continual parameter-efficient fine-tuning improves reliability at the cost of training, maintenance, and potential forgetting. We identify a critical Lazy Agent failure mode where tool necessity is nearly perfectly decodable from mid-layer activations, yet the model remains conservative in entering tool mode, revealing a representation-behavior gap. We propose Activation Steering Adapter (ASA), a training-free, inference-time controller that performs a single-shot mid-layer intervention and targets tool domains via a router-conditioned mixture of steering vectors with a probe-guided signed gate to amplify true intent while suppressing spurious triggers. On MTU-Bench with Qwen2.5-1.5B, ASA improves strict tool-use F1 from 0.18 to 0.50 while reducing the false positive rate from 0.15 to 0.05, using only about 20KB of portable assets and no weight updates.

翻译：使大型语言模型智能体适应特定领域的工具调用，在接口不断演变的背景下仍然非常脆弱。提示词与模式工程易于部署，但在分布偏移和严格解析器下往往表现脆弱；而持续的参数高效微调虽能提升可靠性，却需付出训练、维护及潜在遗忘的代价。我们识别出一种关键的"惰性智能体"失效模式：工具必要性几乎可从中间层激活中完美解码，但模型在进入工具模式时仍趋于保守，这揭示了表征与行为间的差距。我们提出激活导向适配器，这是一种免训练的推理时控制器，它执行单次中间层干预，并通过路由器调制的导向向量混合体（辅以探针引导的符号门）来瞄准工具领域，从而放大真实意图并抑制虚假触发。在MTU-Bench基准测试中，基于Qwen2.5-1.5B模型，ASA将严格工具使用的F1分数从0.18提升至0.50，同时将误报率从0.15降至0.05，且仅需约20KB的可移植资源且无需权重更新。