Evoflux introduces an innovative method for improving the execution of compact language model workflows. By evolving workflow graphs through structured edits and execution feedback, it significantly enhances execution feasibility. This advancement is crucial for optimizing tool agents in real-time applications, potentially reducing costs and deployment risks. The method outperforms existing approaches, indicating a shift towards more reliable execution strategies in AI tool use.