Dashboard

The interactive TUI dashboard is the default way to explore your AI spend.

Launch

codeburn                        # default: last 7 days
codeburn today                  # today's usage
codeburn month                  # this month's usage

Dashboard Panels

Overview - total cost, calls, sessions, cache hit %
Daily chart - cost per day over the selected period
Projects - per-project breakdown with average cost per session
Models - per-model token counts and cost
Activities - per-task-type breakdown with one-shot rate
Core tools - tool usage frequency
Shell commands - most-used bash commands
MCP servers - MCP tool usage
Top sessions - five most expensive sessions across all projects

Keyboard Shortcuts

Key	Action
`1`-`5`	Switch period (Today, 7d, 30d, Month, 6m)
`p`	Toggle provider
`c`	Open model comparison
`o`	Open optimize findings
`b`	Go back
`q`	Quit
Arrow keys	Switch time periods

Auto-Refresh

The dashboard auto-refreshes every 30 seconds by default. Customize or disable:

codeburn report --refresh 60    # refresh every 60 seconds
codeburn report --refresh 0     # disable auto-refresh

Reading the Dashboard

Signal	What it might mean
Cache hit < 80%	System prompt or context is not stable
Lots of `Read` calls	Agent re-reading same files
Low 1-shot rate (30%)	Agent struggling with edits, retry loops
Opus dominating cost	Overpowered model for simple tasks
No MCP usage shown	Not using MCP, or config is broken
`git status` heavy	Agent exploring instead of executing
Conversation dominant	Agent talking instead of doing

These are starting points, not verdicts. A 60% cache hit on a single experimental session is fine. A persistent 60% across weeks is a config issue.

Installation Reports