FlowMind: Execute-Summarize for Structured Workflow Generation from LLM Reasoning

TL;DR

FlowMind introduces an Execute-Summarize framework that separates task execution from workflow construction, improving accuracy and robustness in translating LLM reasoning into structured workflows, validated by extensive experiments on FlowBench.

Contribution

The paper presents a novel decoupled approach for workflow generation from LLM reasoning, outperforming prior integrated methods in accuracy and robustness.

Findings

Outperforms existing methods on FlowBench benchmark.

Decoupling execution and summarization improves workflow accuracy.

Framework enhances robustness in structured workflow generation.

Abstract

LLMs can solve complex tasks through reasoning and tool use, but accurately translating these solutions into structured workflows remains challenging. We model workflows as sequences of tool use and reformulate the problem as designing a mechanism that can both solve tasks and reliably construct workflows. Prior approaches that build workflows during execution often suffer from inaccuracies due to interference between the two processes. We propose an Execute-Summarize(ES) framework that decouples task execution from workflow construction: the model first completes the task using available tools, then independently reconstructs a structured workflow from execution traces. This separation improves workflow accuracy and robustness. We introduce FlowBench and show through extensive experiments that our approach outperforms existing methods, providing a reliable paradigm for grounding free-form LLM reasoning into structured workflows.