Loading request...
When using multi-tool agents with tools like TavilySearch and YFinance, the agent often mixes up which tool to call. Additionally, the results from some tools frequently do not contribute to the final LLM answer, indicating a need for better integration and decision-making logic within the agent's tool-calling mechanism.
I find in multi tool agents, they mix up the tools. For example, when i give tavilysearch and yfinance tools for generic info and finance metrics repectively, it mixes up calling the other . And most of the times, few tools' result never contributes to the final LLM answer..why?