workflow-generation

GrassLobster is an experimental AI-powered tool that generates parametric geometry workflows for Rhino and Grasshopper, enabling users to describe design ideas and create adjustable CAD models.

FlowScout is a framework that automatically generates tool-integrated agentic workflows from historical task-solving records, using Monte Carlo tree search guided by execution feedback. Experiments show it improves tool invocation correctness and execution score over baselines.

The paper introduces a knowledge-centric framework for generating ComfyUI workflows by distilling hierarchical knowledge (pseudo-codes, skeletons, strategies) from real workflows and using LLMs to perform reasoning from task descriptions to executable structures, achieving higher node diversity and execution success rates.

Introduces MetaFlow, a method that trains large language models to generate zero-shot workflows for tasks by combining supervised fine-tuning and reinforcement learning with execution feedback, achieving strong generalization to untrained tasks and operator sets.

A tweet discusses a mental model where an AGI agent functions as a just-in-time workflow generator and executor, noting that Anthropic's Dynamic Workflows are an early implementation of this concept.