aqua5230/usage — MacOS menu bar app pinning Claude Code & Codex quota, tokens, and cost to your screen.

Description and use

**Summary**
aqua5230/usage is a lightweight macOS menu‑bar utility that displays Claude Code & Codex usage metrics—quota, token consumption, and estimated cost—directly on your screen. It runs entirely locally (no API calls), offers HTML reporting and nine visual themes, and is written in Python. With 224 ★ and recent commits, it’s a well‑maintained OSS tool for developers who need instant visibility into their AI usage.

**Value**
The app gives developers immediate feedback on how much of their Claude or Codex budget is being spent, helping to curb unexpected costs during prototyping, RAG pipelines, or agent‑based workflows. Because it’s local‑only, there’s no extra latency or network‑security surface, and the built‑in HTML reports make it easy to audit usage over time or share snapshots with stakeholders.

**Practical adoption path**

1. **Install** the Python package (or download the pre‑built binary) and add the app to the macOS menu bar.
2. **Configure** your Claude Code / Codex credentials once; the tool will automatically read token usage from the local SDK/CLI logs.
3. **Select a theme** and enable HTML reporting if you need periodic usage summaries.
4. **Integrate** the exposed signals (API/SDK/CLI usage data) into your CI/CD or monitoring dashboards, or use the generated reports to fine‑tune prompts and model calls during development.

**Production readiness**
The project shows strong production signals: recent activity (last commit 2026‑06‑23), solid community interest (224 ★, 39 forks), a clear Python codebase, and a well‑defined feature set. While the license and security posture still need a final check, the overall health, documentation, and low‑risk architecture (local execution, no external calls) make it ready for a serious pilot in internal tools or developer workstations.

Problem

aqua5230/usage helps add AI capability without starting from a blank model stack.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• prototype AI features
• build RAG or agent workflows
• evaluate model tooling

Evidence

macOS menu bar app pinning Claude Code & Codex quota, tokens, and cost to your screen. Local-only, zero API calls. HTML reports, 9 themes.

Source