MermaidFlow: Redefining Agentic Workflow Generation via Safety-Constrained Evolutionary Programming

TL;DR

MermaidFlow introduces a safety-constrained, graph-based evolutionary framework for generating robust, verifiable workflows in autonomous agent reasoning, significantly improving success rates and convergence speed without altering task protocols.

Contribution

It redefines workflow generation by integrating safety constraints and structured graph evolution, enhancing robustness and interpretability in agentic reasoning systems.

Findings

Improves success rates of workflow generation

Achieves faster convergence to executable plans

Provides a scalable, modular framework for agent reasoning

Abstract

Despite the promise of autonomous agentic reasoning, existing workflow generation methods frequently produce fragile, unexecutable plans due to unconstrained LLM-driven construction. We introduce MermaidFlow, a framework that redefines the agentic search space through safety-constrained graph evolution. At its core, MermaidFlow represent workflows as a verifiable intermediate representation using Mermaid, a structured and human-interpretable graph language. We formulate domain-aware evolutionary operators, i.e., crossover, mutation, insertion, and deletion, to preserve semantic correctness while promoting structural diversity, enabling efficient exploration of a high-quality, statically verifiable workflow space. Without modifying task settings or evaluation protocols, MermaidFlow achieves consistent improvements in success rates and faster convergence to executable plans on the agent reasoning benchmark. The experimental results demonstrate that safety-constrained graph evolution offers a scalable, modular foundation for robust and interpretable agentic reasoning systems.