See where your AI spend goes — by task, model, tool, and project, across 31 AI tools (Claude Code, Codex, Cursor, Gemini CLI, Cline, and more).
When to Use
- "How much am I spending on Claude Code?"
- "Show me my token usage breakdown"
- "Which tasks are costing the most?"
- "Am I using the right model for the job?"
- "Optimize my AI spend"
- Any cost/budget/token-usage question about AI tools
Quick Start
# Interactive dashboard (last 7 days)
npx codeburn
npm install -g codeburn codeburn
## Core Commands
```bash
# Dashboard views
codeburn # Interactive TUI — last 7 days (default)
codeburn today # Today's usage
codeburn month # This calendar month
# One-liner status
codeburn status # Today + month totals
codeburn status --format json # Same as JSON
# Reports
codeburn overview # Plain-text monthly summary
codeburn report -p 30days # Rolling 30-day window
codeburn report -p all # All recorded sessions
codeburn report --from 2026-04-01 --to 2026-04-30 # Date range
codeburn report --format json # Full data as JSON
# Export
codeburn export # CSV (today, 7d, 30d)
codeburn export -f json # JSON export
# Analysis
codeburn optimize # Scan for waste + copy-paste fixes (30d)
codeburn optimize -p week # Scope to last 7 days
codeburn compare # Side-by-side model comparison
# Web dashboard with charts
codeburn web # http://localhost:4747
# Filter by provider or project
codeburn report --provider claude
codeburn report --project my-app
Task Categories (Auto-Classified)
| Category | What triggers it |
|---|---|
| Coding | Edit, Write tools |
| Debugging | Error/fix keywords + tool usage |
| Feature Dev | "add", "create", "implement" |
| Refactoring | "refactor", "rename", "simplify" |
| Testing | pytest, vitest, jest in Bash |
| Exploration | Read, Grep, WebSearch without edits |
| Planning | EnterPlanMode, TaskCreate tools |
| Delegation | Agent tool spawns |
| Git Ops | git push/commit/merge in Bash |
| Build/Deploy | npm build, docker, pm2 |
| Brainstorming | "brainstorm", "what if", "design" |
| Conversation | No tools, pure text exchange |
| General | Skill tool, uncategorized |
Reading the Dashboard — Key Signals
| Signal | What it might mean |
|---|---|
| Cache hit < 80% | System prompt or context not stable |
Lots of Read calls per session |
Agent re-reading same files, missing context |
| Low 1-shot rate (Coding 30%) | Agent struggling with edits, retry loops |
| Opus dominating cost on small turns | Overpowered model for simple tasks |
| Conversation category dominant | Agent talking instead of doing |
| No MCP usage shown | Either not using MCP, or config is broken |
Claude Code Data Location
CodeBurn reads Claude Code session files from:
~/.claude/projects/<sanitized-path>/<session-id>.jsonl
Each session carries: model name, token usage (input/output/cache read/cache write), tool_use blocks, timestamps. No data leaves your machine.
Multiple Claude config dirs: set CLAUDE_CONFIG_DIRS (: on POSIX, ; on Windows).
Installation
Already installed in this project. On fresh clones, setup.sh installs via:
npm install -g codeburn
Verify:
codeburn --version
Upgrade
npm install -g codeburn@latest
# Or via /update-all skill