AI coding agent actions visualized as timelines for easier understanding
Agent ATO: Visualizing Agent Interaction Timelines from Logs
Software Engineering
Summary
AI coding helpers do many things while fixing or writing code, like searching files, reading, editing, and testing, but it’s hard to see what they did just by looking at the final code. The authors created a tool called Agent ATO that turns the AI’s recorded actions from console logs into easy-to-read timelines. These timelines show different parts of the agent’s work, such as finding files or running tests, helping developers understand and compare what the AI did. This can make it easier to check how AI helpers work during coding tasks.
AI coding agentssoftware repositoriesconsole logscode editingtimeline visualizationfile discoveryprogram testingdeveloper tools
Authors
Takuto Kawamoto, Yoshiki Hig, Raula Gaikovina Kula
Abstract
AI coding agents are becoming part of developers' workflows, but their behavior is difficult to understand from final code changes alone. During a task, agents interact with software repositories through sequences of actions such as searching for files, reading code, editing programs, and running tests or build commands. These interactions, together with token usage, are often recorded in console logs, but raw logs are difficult for developers to inspect. In this paper, we propose Agent ATO (Agentic Trajectory Observer), a tool for visualizing AI coding agent interaction timelines from console logs. Agent ATO extracts agent interactions, classifies them by command or tool type, and visualizes them as timelines. In addition to an all-interaction timeline, Agent ATO provides filtered timelines that emphasize file discovery, file reading, file editing, and execution while preserving surrounding context. We illustrate how Agent ATO may help developers inspect and compare agent actions using selected runs from two repair tasks. Future work will apply Agent ATO to more agents, tasks, and development environments, and will evaluate whether it reduces the effort needed to compare trajectories.