Don’t worry, as agents are on the rise, LLMs are learning iterative testing & tool use. I just published a new MCP tool for Cline which uses WolframAlpha to solve maths and other knowledge. It works great, and Cline can iterate to try different variants of prompts it sends to Wolfram if the syntax isn’t quite right. Then it ingests the result into context and can continue with the knowledge it gained
Don’t worry, as agents are on the rise, LLMs are learning iterative testing & tool use. I just published a new MCP tool for Cline which uses WolframAlpha to solve maths and other knowledge. It works great, and Cline can iterate to try different variants of prompts it sends to Wolfram if the syntax isn’t quite right. Then it ingests the result into context and can continue with the knowledge it gained