Dashboard

The dashboard provides a visual overview of your AI coding activity. It aggregates data from all scanned sessions to surface trends and insights that complement the session browser and search workflow.

Stats Overview

The top row shows key metrics at a glance:

• Total Sessions — number of sessions across all scanned directories
• Total Messages — combined user + assistant message count
• User Messages — messages you sent
• Assistant Messages — AI responses
• Average Messages/Session — conversation depth indicator

A weekly comparison badge shows the trend vs. the previous week.

Activity Heatmap

A GitHub-style contribution heatmap showing daily session activity over the past year. Darker cells indicate more sessions on that day. Hover for exact counts.

Project Distribution

A chart showing which projects (working directories) have the most sessions. Useful for understanding where you spend your AI-assisted development time.

Model Usage

Breakdown of which AI models were used across sessions. Tracks model switches within sessions (Pi logs model_change entries in JSONL).

Token Costs

Aggregated token usage and estimated costs across sessions. Helps track spending on AI-assisted development.

Achievements

A gamification layer that awards badges based on your usage patterns:

• Session milestones (10, 50, 100, 500 sessions)
• Message volume achievements
• Streak tracking (consecutive days with sessions)
• Project diversity badges

Top Projects

A ranked list of your most active projects by session count, with quick navigation to filter the session browser by project.

Performance

The dashboard uses incremental data aggregation — stats are computed from the cached session metadata rather than re-parsing JSONL files. This keeps the dashboard responsive even with thousands of sessions.