Post-task workflows improve understanding and reuse of AI agent tasks
From Review to Reuse: How Post-Task Workflow Can Support Human-AI Agent Interaction
Human-Computer Interaction
Summary
When AI agents complete tasks, people often only see the final result and some scattered details, making it hard to understand how the AI did the job or to fix mistakes. The authors studied a way to show the full process after the task finishes, using clear, editable diagrams called post-task workflows. They found that seeing these workflows helped people better spot errors and understand the AI's work, and that editing these workflows was as effective and sometimes preferred over just changing the original request. This approach makes it easier to check and reuse AI work.
What this means in practice
- •For workflow automation developers: Provide interactive visual summaries of AI task executions to help users verify and edit automated workflows more effectively.
- •For customer support teams: Use post-task workflows to better understand AI-driven automation processes when troubleshooting or updating support tools.
Authors
Zekun Wu, Xinru Wang, Rock Yuren Pang, Chenglong Wang, Anna Maria Feit
Abstract
AI agents can automate tasks by turning a single natural-language request into a multi-step process spanning tools, files, and applications. Users are often left to judge that process from fragmented execution information and the final output. To make the completed process easier to understand, validate, and reuse, we investigate post-task workflows: editable, graph-based representations of an agent's completed execution. We first analyzed 10,803 public workflow templates from n8n to characterize real-world automation practice, then developed Trace2Flow, a research probe that translates agent execution traces into interactive post-task workflows. In a study, participants (N = 20) reviewed agent executions with prompt or agent errors. We found that post-task workflows improved their understanding and error detection over a prompt-only condition, and that validation succeeded mainly when users cross-checked across multiple evidence sources. For follow-up tasks, adapting the workflow matched adapting the prior prompt in success, time, and difficulty, and was often preferred.