@851-labs/tokenmaxxing

CLI for tokenmaxxing — the social leaderboard for LLM token usage. Parses your local agent usage (Claude Code, Codex, OpenCode, Gemini CLI, Copilot CLI, Hermes, Pi, Oh My Pi, Grok Build CLI, Antigravity, ZCode, Amp, Qwen Code, Kimi CLI, Kilo Code, Goose, Droid, Codebuff, OpenClaw) via ccusage and pushes daily aggregates to your public profile.

Usage

npm install -g @851-labs/tokenmaxxing@latest
tokenmaxxing login              # sign in in the browser, approves this device
tokenmaxxing sync               # parse local usage and push it
tokenmaxxing service install    # optional: sync automatically every 5 minutes
tokenmaxxing upgrade            # upgrade the global CLI and refresh the service

You can also install globally with bun add -g --trust @851-labs/tokenmaxxing@latest, pnpm add -g @851-labs/tokenmaxxing@latest, or yarn global add @851-labs/tokenmaxxing@latest.

The background service uses the global tokenmaxxing binary and syncs every 5 minutes. It auto-updates through the package manager that installed the global binary (bun, npm, pnpm, or yarn) when that package manager can be detected. Custom log roots (CLAUDE_CONFIG_DIR, CODEX_HOME, HERMES_HOME, the other agents' data-directory variables such as OPENCODE_DATA_DIR, GEMINI_DATA_DIR, COPILOT_HOME, PI_AGENT_DIR, GROK_HOME or AMP_DATA_DIR, and XDG_DATA_HOME/XDG_CONFIG_HOME) are captured at tokenmaxxing service install or tokenmaxxing service repair; rerun one of those after changing them. Without HERMES_HOME, Hermes usage includes every named profile under ~/.hermes/profiles/. Use tokenmaxxing service status for the last run and tokenmaxxing service doctor to inspect scheduler files, auth, auto-update, locks, and recent logs. Each doctor check is OK, INFO (fine, but worth knowing: never synced yet, a sync running right now), WARN (syncs still run, but something is off) or FAIL (scheduled syncs can't happen); every WARN and FAIL line ends with the command that fixes it. service doctor exits 1 when any check is WARN or FAIL and 0 otherwise, so scripts and CI can gate on it. With --json, the report's health is ok, warn or fail (status only says the doctor ran).

Run sync as often as you like, from as many machines as you like — profiles aggregate across devices. Useful flags: --dry-run, --since YYYY-MM-DD, --sources claude,codex, --json. sync exits non-zero when no agent's usage could be collected; if only some agents fail, it still uploads the rest and exits 0 with "status": "partial" in --json output.

What gets uploaded (privacy)

Daily aggregates only: date, model name, agent name, token counts, and the API-equivalent cost — never prompts, file paths, project names, or session content. Revoke access anytime with tokenmaxxing logout or from settings.

License

MIT