Last updated 29 July 2026
This app has no servers and collects nothing. Everything happens on your Mac, and the only services it talks to are Anthropic and OpenAI, reusing the Claude Code and ChatGPT/Codex logins you already have.
This policy covers the AI Usage Monitor macOS app (“the app”),
distributed at
github.com/stavrop/ai-usage-monitor
and via the stavrop/tap Homebrew tap.
Nothing — for the developer. There is no server, no analytics, no telemetry, no crash reporting, no tracking, and no account with the developer. The developer never receives your credentials or your usage data. The app runs entirely on your Mac and communicates only with Anthropic's and OpenAI's servers to read your usage.
The app uses only logins that are already on your Mac. It never asks you to sign in, and only ever contacts each provider with that provider's own token. If a provider isn't signed in, the app doesn't show it and makes no requests to it.
Claude Code-credentials).
It refreshes the access token when it expires and writes it back to that same
item.api.anthropic.com, platform.claude.com) to read your
usage — never anywhere else, and never to the developer.~/.codex/auth.json. This is
read-only: the app never writes to that file and never refreshes
that token, because those tools own it. An expired token is reported in the menu
instead.chatgpt.com) to read your usage — never anywhere else, and never
to the developer.~/.codex/auth.json and
belongs to those tools; the app only reads it and keeps no copy.The app is not directed at children and collects no personal information from anyone.
This policy may change; updates are published here with a new “last updated” date.
Questions or concerns: open an issue at github.com/stavrop/ai-usage-monitor/issues.