dashboard

To visualize agent efficiency metrics from local logs, you can parse the results_*.json and summary_*.md files stored in the .claude/.agent_outputs/ directory to generate markdown dashboards and summaries.

Yes, you can generate a metrics dashboard for task completion rates and error patterns by aggregating local agent logs to reveal performance trends over configurable time ranges.

You can track consensus scores and model usage across multiple agents by running a skill that reads local agent output directories and aggregates the parsed data into markdown summaries.

The best way to prepare log data for a sprint review or capacity planning is to aggregate task completion rates and model usage from local logs into a visual markdown dashboard, helping you quickly identify performance bottlenecks.

No, the dashboard skill requires no external dependencies to parse agent logs, as it reads local .claude/.agent_outputs/ data directly and applies safety checks during the parsing and aggregation process.