llm 0.37.2

llm 0.37.2 was released to fix a security issue reported by Daniel de Araujo in GHSA-v9h7-jj4g-f862 and GHSA-r639-rr7r-x72w. The fix involves managing the extra-openai-models.yaml mechanism via the new-in-0.37 llm openai endpoint add command. Daniel reported two ways the configured OpenAI API key could be accidentally sent to other providers.