Worked on the langchain-ai/langchain-google repository to address a critical bug affecting numeric argument handling in tool calls. Focused on the integration between Python and Protobuf, the developer implemented a fix to ensure that integer-like float values are correctly preserved as integers when processed through tool call arguments. This change stabilized function-call behavior and maintained data integrity across tool invocations. To validate the solution, a robust unit test was added, verifying accurate int and float type conversions. The work demonstrated attention to detail in data type conversion, tool calling, and unit testing, improving the reliability of generative AI tool pipelines.