llm 0.31.1 release: bug fix for OpenAI Chat Completion tool calls

The open-source llm project fixes a specific bug in version 0.31.1. The update resolves an issue where a tool call with empty arguments could produce a JSON error with some providers when using OpenAI Chat Completion endpoints. The fix is related to issue number 1521 (#1521). The bug appeared when developers tested with the llm-meta-ai component and reported it. The release aims to improve compatibility and stability when integrating with various providers, reducing the risk of processing errors or crashes caused by empty arguments, and making Chat Completion–based workflows with third-party providers more reliable. The update is important for developers and administrators integrating OpenAI or other providers into llm. The short release description mentions only the bug fix and contains no additional feature enhancements; the 0.31.1 version number indicates a minor, bug-fix release.