AGILE: A Novel Reinforcement Learning Framework of LLM Agents | Read Paper on Bytez