How can I view my Claude Code usage?

Install the CLI and run crbon install. It hooks Claude Code and records every session’s tokens per model on your machine, and the dashboard’s My coding agents view shows them per agent, per model and per week with cost and CO2e. Your plan’s usage limits stay in Anthropic’s own app; the CLI reports what you used, not what you have left.

How can I check my Codex usage?

The same way. The CLI hooks Codex alongside Claude Code, Gemini CLI, Cursor and OpenClaw, so its sessions land in the same view with the same per-model breakdown. OpenAI’s plan limits and credits are shown in ChatGPT and on the OpenAI platform, not in CrbonFree.

How often does it sync?

The job that crbon install registers with your operating system runs every hour and uploads whatever is new. If you want the dashboard current right now, run crbon sync.

The CLI is free and MIT licensed, and tracking usage and carbon estimates on the free plan costs nothing. Signed receipts, audit packs and trust assets are part of the paid plans.

Which coding agents does it meter?

Claude Code, Codex, Gemini CLI, Cursor and OpenClaw, each through a dedicated hook the installer wires up. A local proxy can also capture direct API traffic for tools without a hook.

Does it read my prompts?

Metering needs only usage metadata: model, tokens and timestamps. The optional shared-memory tools store redacted session content so agents can search past work, and redaction, sensitive-folder exclusion and .env blocking run before anything leaves your machine.

Can I run it on a server or in CI?

Yes. On a machine without a browser, crbon install shows a device code you approve from the dashboard. You can also set CRBON_API_KEY in the environment and skip the sign-in entirely.

@crbonfree/cliv0.1.11Measure what you use

Your Claude Code and Codex usage, metered from their own logs.

One install hooks the coding agents you already run, records their token usage locally and syncs it to your dashboard on a schedule, per agent, per model and per week. The backend adds the cost and the carbon.

Install and sign in

npm install -g @crbonfree/cli && crbon install

View on npm See it in the sample account

illustrative session

4,500+

SDK and MCP downloads with no paid marketing

Crbon Labs, 2026

coding agents metered through their own hooks

MCP tools for shared agent memory

layers in the carbon methodology, version 1.2

What is the crbon CLI?

The crbon CLI is a command-line usage tracker for AI coding agents. Installed from npm, it hooks Claude Code, Codex, Gemini CLI, Cursor and OpenClaw, records their token usage on your machine and syncs it to your CrbonFree dashboard on a schedule, per agent, per model and per week, with cost and CO2e from the published methodology. It also ships a small MCP server for shared agent memory.

Last checked against the product on 5 September 2026. Carbon figures follow methodology v1.2.

At a glance

Package: @crbonfree/cli on npm
Version: 0.1.11
Licence: MIT
Requires: Node.js 20 or newer
Runs on: macOS and Linux, via launchd or systemd
Meters: Claude Code, Codex, Gemini CLI, Cursor, OpenClaw
Data direction: Captures locally, uploads on a schedule
Signs in with: crbon install, or CRBON_API_KEY

Hooks for the agents you already run

Claude Code, Codex, Gemini CLI, Cursor and OpenClaw each get their own hook, so usage is captured per agent and per model as it happens, with nothing to change in how you work.

Local first, then a scheduled sync
Capture lands in a folder on your machine. crbon install registers a job with your operating system’s scheduler, and each sync uploads what is new. The backend turns it into CO2e with the published methodology.
Redaction before anything leaves
Redaction is always on. Sensitive folders are excluded, .env files are blocked and secrets are scrubbed before an upload starts.
A shared memory across agents
The CLI ships its own MCP server with seven tools, so any agent can search past sessions, summarise them, brief itself on a file or hand work to another agent. Search scope is yours, your team’s or the whole organisation’s, as your org settings allow.

How the crbon CLI moves your data

1
An agent runs
Each supported agent gets its own hook, so every call is recorded with its model and token counts as it happens.
2
Capture lands on disk
Sessions are written to a folder on your machine. Redaction, sensitive-folder exclusion and .env blocking run here, before anything leaves.
3
The scheduler syncs
The job crbon install registered uploads what is new every hour. Run crbon sync to push immediately.
4
The dashboard reports
The backend converts tokens to CO2e with methodology v1.2 and My coding agents shows the result per model, per agent and per week.

How to get started

1
Install
Node 20 or newer. The package is MIT licensed.
```
npm install -g @crbonfree/cli
```
2
Sign in and set up
Opens the browser to sign in, provisions an API key, registers the scheduler and runs the first sync. On a headless machine it falls back to a device code you approve from the dashboard.
```
crbon install
```
3
Check what it found
Lists the agents it is metering and the time of the last sync. Run crbon sync to push immediately instead of waiting for the schedule.
```
crbon status
```

It lands in My coding agents

Every synced session shows up in the dashboard per model, per agent and per week, next to connection health and the last sync time. Any member of your organisation can create their own CLI key, so a team can start without an admin.

Open the sample account

What it does not do

It meters the supported agents on the machine it runs on and needs an account and an API key; there is no anonymous mode. The token-saving features, compressed output, bounded reads and loop detection, are available for Claude Code only today. Usage metering needs only metadata, but the shared-memory tools store redacted session content on the server so your agents can search it.

Who reaches for the crbon CLI

Engineering lead

Wants to know what the team’s coding agents cost in tokens and carbon without asking anyone to change tools.

What they get: Every developer runs one install, and the dashboard shows usage per agent, per model and per week.

Sustainability officer

Needs developer AI usage inside the same Scope 3 report as the production workloads.

What they get: CLI sessions land in the same methodology and the same monthly audit pack as provider connections.

Developer

Switches between agents during the day and loses the thread each time.

What they get: The shared-memory tools let any agent search, summarise and hand off past sessions.

Which CrbonFree surface do I need?

Three of them put usage in: the dashboard through provider connections, the CLI for coding agents and the browser extension for chat sessions. The other four read the numbers back into your own code. The crbon CLI is highlighted.

Surface	Meters usage	Reads numbers back	Runs	Needs	Language
Dashboard			In the browser	An account; the sample account opens without one	None
CLIthis page		status	On your machine	An account; device code for headless machines	Node 20 or newer
Browser extension	estimates		In Chrome	Optional, per the store listing	None
TypeScript SDK			Node 18 or newer	An API key	TypeScript or JavaScript
Python SDK			Python 3.8 or newer	An API key	Python
MCP server			Beside your MCP client	An API key from login	Any MCP client
REST API			Over HTTPS	An API key	Any

Questions

Questions about the crbon CLI.

Short answers. The package page and the API reference have the long ones.

View on npm

Install the CLI and run crbon install. It hooks Claude Code and records every session’s tokens per model on your machine, and the dashboard’s My coding agents view shows them per agent, per model and per week with cost and CO2e. Your plan’s usage limits stay in Anthropic’s own app; the CLI reports what you used, not what you have left.
The same way. The CLI hooks Codex alongside Claude Code, Gemini CLI, Cursor and OpenClaw, so its sessions land in the same view with the same per-model breakdown. OpenAI’s plan limits and credits are shown in ChatGPT and on the OpenAI platform, not in CrbonFree.
The job that crbon install registers with your operating system runs every hour and uploads whatever is new. If you want the dashboard current right now, run crbon sync.
The CLI is free and MIT licensed, and tracking usage and carbon estimates on the free plan costs nothing. Signed receipts, audit packs and trust assets are part of the paid plans.
Claude Code, Codex, Gemini CLI, Cursor and OpenClaw, each through a dedicated hook the installer wires up. A local proxy can also capture direct API traffic for tools without a hook.
Metering needs only usage metadata: model, tokens and timestamps. The optional shared-memory tools store redacted session content so agents can search past work, and redaction, sensitive-folder exclusion and .env blocking run before anything leaves your machine.
Yes. On a machine without a browser, crbon install shows a device code you approve from the dashboard. You can also set CRBON_API_KEY in the environment and skip the sign-in entirely.

Goes well with

All products

DashboardEvery provider, project and coding agent in one report. The sample account opens without a login.Browser extensionEstimates token usage in your AI chat sessions across ChatGPT, Gemini and the Claude apps, including Claude Code. Guest mode keeps everything on your machine, and signing in keeps the history.TypeScript SDKA typed client for your usage, emissions, receipts and audit packs from Node.js or TypeScript.

See your own number instead of ours.

However your team already runs AI, that is how you connect. Paste a read-only provider key, drop in an SDK, add the MCP server, install the CLI or switch on the browser extension, and the first measured tokens reach your dashboard within seconds.

Create your free account

Free plan · first tokens in seconds

Your Claude Code and Codex usage, metered from their own logs.

What is the crbon CLI?

Hooks for the agents you already run

Local first, then a scheduled sync

Redaction before anything leaves

A shared memory across agents

How the crbon CLI moves your data

An agent runs

Capture lands on disk

The scheduler syncs

The dashboard reports