| name | toll |
| description | Use this skill when the user asks about token usage, API costs, spending, or session statistics for Claude Code, Codex CLI, Kimi Code, or Gemini CLI. This skill runs the `toll` CLI to aggregate token counts and estimated USD costs from local session logs stored in ~/.claude, ~/.codex, ~/.kimi, and ~/.gemini. Activate when the user says things like "how many tokens have I used", "how much have I spent on Claude", "show my usage", "token stats", "daily usage breakdown", or "export my usage as CSV".
|
| license | MIT |
| compatibility | Requires the `toll` binary to be installed. Supports Linux and macOS (x86_64 and ARM64). Reads logs from ~/.claude/projects/**/*.jsonl (Claude Code), ~/.codex/sessions/ (Codex CLI), ~/.kimi/sessions/ (Kimi Code), and ~/.gemini/tmp/ (Gemini CLI). No network access required at runtime. |
| metadata | {"author":"Fullstop000","version":"1.0.7","repository":"https://github.com/Fullstop000/toll","homepage":"https://github.com/Fullstop000/toll"} |
toll — Token Usage Tracker
Track token consumption and estimated USD costs from Claude Code, Codex CLI, Kimi Code, and Gemini CLI sessions.
Installation
Check if toll is installed:
toll --version
If not found, install it:
curl -fsSL https://raw.githubusercontent.com/Fullstop000/toll/refs/heads/master/install.sh | sh
cargo install toll
Commands
Map the user's intent to the appropriate toll invocation:
| User Intent | Command |
|---|
| All-time summary | toll |
| Today only | toll --today |
| Last N days | toll --days <N> |
| Daily breakdown | toll --by-day |
| Daily breakdown for N days | toll --by-day --days <N> |
| Claude Code only | toll --claude |
| Codex CLI only | toll --codex |
| Kimi Code only | toll --kimi |
| Gemini CLI only | toll --gemini |
| Combined filters | toll --days <N> --by-day --detail --json |
| Full token counts (no abbreviation) | toll --detail |
| JSON output | toll --json |
| CSV export | toll --csv |
| List model prices | toll --list-prices |
Flags can be combined freely, e.g. toll --by-day --days 7 --claude --detail.
Output Metrics
Explain these columns if the user is unfamiliar:
- Sessions — number of AI coding sessions tracked
- Input — total tokens sent to the model (prompt)
- Cached — tokens served from prompt cache (cheaper rate)
- Hit Rate — percentage of input tokens that were cached
- Net Input — non-cached input tokens billed at full rate
- Output — tokens generated by the model
- Cost — estimated USD cost based on per-model pricing (marked
* when approximate)
Token counts use compact notation: 1.2m = 1,200,000 tokens, 45.6k = 45,600 tokens.
