Language hooks: a modular framework for augmenting LLM reasoning that decouples tool usage from the model and its prompt

Open in new window