Show HN: What's Agent Doing – a Claude Code UI mod that explains each step (github.com)

A new plugin called "What's Agent Doing" has been launched for Claude Code, allowing users to see real-time updates of Claude's actions in plain English. This feature is particularly significant as it enhances transparency in the AI's decision-making process, making it easier for developers to understand what the AI is working on, whether it is stuck, and the purpose behind each command it executes. For example, when fixing a cart total issue, users can observe Claude’s workflow, including commands related to testing and debugging, along with the time each step takes. This plugin is pivotal for the AI/ML community as it addresses the common frustrations of working with complex models that act behind the scenes. By providing a detailed history and status updates of ongoing tasks, it allows developers to more effectively manage and troubleshoot AI functionalities. Built using Claude Code's function hooks, the plugin not only improves user interaction with AI systems but also sets a precedent for future enhancements that could include more detailed insights and user control without affecting the AI's operational integrity. The "What's Agent Doing" plugin is currently available for free, further encouraging community engagement and experimentation.