setup, delegated

Let your agent put you
on the leaderboard.

Copy the prompt, paste it into whatever you code with, and it does the rest: finds your agents, reads the token counts already on your disk, shows you the numbers, and only then offers to sign you in.

paste into Claude Code

Set up burnlog for me — the public leaderboard for AI coding token usage.

Fetch https://burnlog.net/agent-setup.md and follow it top to bottom. It is written to be executed, not summarised.

Rules: show me my own totals before any account exists, ask me before anything is published, and never upload prompts, code, file names, or repo names — token counts only.

I use Claude Code, which writes its own usage log, so the plain sync path covers it.

your agent asks before it publishes anythingread what it will run →

What your agent will do

1Read what's already on your diskClaude Code and Codex write their own session logs. Your first sync counts history you accumulated before you had heard of burnlog — nothing to set up retroactively.
2Show you the number before asking for anythingThe scan runs signed out and uploads nothing. The agent is instructed to report your totals back to you first, because an account you agreed to after seeing the number is a different decision from one you agreed to before.
3Hand you the GitHub sign-inIt opens the browser and steps back. The instructions forbid it from automating the login or touching your credentials; with no browser available it tells you to run the login yourself.
4Install the session-end hookSo every future session syncs itself. It tells you which config file it changed rather than editing it quietly.
5Cover the agents with no usage logCursor on your own key, Gemini CLI, aider, opencode, your own scripts: burnlog runs them under a loopback proxy and reads the provider's own usage field. No certificate to install, and the request body is never inspected.

Rather do it yourself

The prompt is a convenience, not a requirement. The same setup is one command:

$ npx @sxnalabs/burnlog

the instructions your agent follows: /agent-setup.md · what is stored: /privacy · per-agent detail: /tools

Let your agent put you
on the leaderboard.

paste into Claude Code

Set up burnlog for me — the public leaderboard for AI coding token usage.

Fetch https://burnlog.net/agent-setup.md and follow it top to bottom. It is written to be executed, not summarised.

Rules: show me my own totals before any account exists, ask me before anything is published, and never upload prompts, code, file names, or repo names — token counts only.

I use Claude Code, which writes its own usage log, so the plain sync path covers it.

your agent asks before it publishes anythingread what it will run →

What your agent will do

1Read what's already on your diskClaude Code and Codex write their own session logs. Your first sync counts history you accumulated before you had heard of burnlog — nothing to set up retroactively.
2Show you the number before asking for anythingThe scan runs signed out and uploads nothing. The agent is instructed to report your totals back to you first, because an account you agreed to after seeing the number is a different decision from one you agreed to before.
3Hand you the GitHub sign-inIt opens the browser and steps back. The instructions forbid it from automating the login or touching your credentials; with no browser available it tells you to run the login yourself.
4Install the session-end hookSo every future session syncs itself. It tells you which config file it changed rather than editing it quietly.
5Cover the agents with no usage logCursor on your own key, Gemini CLI, aider, opencode, your own scripts: burnlog runs them under a loopback proxy and reads the provider's own usage field. No certificate to install, and the request body is never inspected.

Rather do it yourself

The prompt is a convenience, not a requirement. The same setup is one command:

$ npx @sxnalabs/burnlog

the instructions your agent follows: /agent-setup.md · what is stored: /privacy · per-agent detail: /tools

Let your agent put youon the leaderboard.

What your agent will do

Rather do it yourself

Let your agent put youon the leaderboard.

What your agent will do

Rather do it yourself

Let your agent put you
on the leaderboard.

Let your agent put you
on the leaderboard.