You can now hear your AI agents work in Claude Code.

Follow agent execution in real time, understand decisions as they happen, and debug workflows without guesswork.

Key takeaways

• AI agents don’t have to run silently, you can make their execution observable
• Hearing agent activity helps you understand decisions as they happen
• Better observability leads to faster debugging and iteration
• Works naturally with Claude Code way of agent workflows, make BMAD agents talk
• Turns “trust the output” into reason about the process

Full implementation details: Hear Your AI Agents Work in Claude Code