What problem does it solve?
This Skill provides a real-time, visual dashboard to monitor Claude Code agent interactions, making it easy to understand complex multi-agent workflows, debug issues, and track performance. It transforms opaque agent execution into transparent, actionable insights without needing a separate database.
Core Features & Use Cases
- Real-time Event Streaming: Visualizes agent activities, tool usage, and session flows as they happen, providing immediate feedback on AI operations.
- Agent Swim Lanes: Clearly displays parallel execution of multiple agents, enhancing understanding of complex orchestrations and inter-agent communication.
- Filesystem-Based: Leverages Claude Code's hooks to capture events to JSONL files, eliminating external database dependencies and simplifying setup and maintenance.
- Use Case: A developer is building a multi-agent system where a "Planner" agent launches "Researcher" and "Writer" sub-agents. This dashboard allows them to see exactly when each agent starts, which tools they use, and how long each step takes, helping them identify bottlenecks or unexpected behaviors for rapid iteration.
Quick Start
Start the agent observability dashboard.