ccr — Claude Code run-rate

Know how fast you're burning your Claude Code subscription window — and whether clearing context now is worth it — without doing token math in your head.

A solo side project, shared as-is — built to scratch my own itch and daily-driven. See Status & support for what to expect, and the acceptance criteria in features/ for exactly how the burn-rate and clear-ROI math behaves.

---

What it does

ccr reads Claude Code's own status-line JSON and your session transcripts and shows you the economy of a session:

• Burn rate on your 5-hour (and weekly) rate-limit window — %/min and, more usefully, how long until you hit the wall.
• Clear ROI — in plain language, how many more minutes clearing context now would buy you.
• Live tool/skills feed — in the sidebar, a rolling view of recent tool and slash-command calls plus per-session stats (files touched, tokens generated).
• Resume advisor (ccr resume) — recent sessions ranked by what they'd cost to bring back (context size, share of the window, cold/warm cache), then it hands selection to claude --resume.

For scripting and external tools (status bars, menu-bar widgets), ccr economy --json emits a stable, versioned model — see docs/JSON-CONTRACT.md.

It is read-only: it never modifies ~/.claude/settings.json, your credentials, or any config. It targets subscription plans (Pro / Max); on API/pay-as-you-go it degrades gracefully and just shows Claude Code's own cost.

Quickstart

# zero-install, on-demand report (any OS):
npx claude-code-runrate economy

# or install the `ccr` command for the live statusline / sidebar:
npm i -g claude-code-runrate

Surfaces

Surface	What	Linux	macOS	Windows
ccr economy (CLI report)	Full panel on demand	✅	✅	✅
ccr economy --json	Stable machine-readable model (scripting)	✅	✅	✅
ccr resume	Recent sessions ranked by cost to resume	✅	✅	✅
ccr statusline	One-line summary wired into CC's statusLine	✅	✅	✅
ccr sidecar	Live dashboard + tool/skills feed	✅ tmux	✅ tmux	✅ Windows Terminal / VS Code

The CLI and statusline are pure Node — they run on native Windows. The live sidebar runs on Windows Terminal (ccr splits it automatically) or inside VS Code's integrated terminal on any OS (see VS Code). On Linux/macOS the default host is tmux. No WSL required.

Requirements

• Node ≥ 18.3 — that's it for the core. Zero runtime dependencies.
• For the optional live ccr sidecar: tmux on Linux/macOS, or Windows Terminal / VS Code's integrated terminal on Windows (no WSL).

VS Code (split terminal)

ccr detects VS Code's integrated terminal (TERM_PROGRAM=vscode) and wires the live sidebar into a split pane — no separate window, no WSL. A shell can't trigger the split itself, so ccr does everything around it:

1. Run ccr (or ccr <profile>). Claude starts in the current pane and a bright banner shows the steps. The sidecar command is copied to your clipboard automatically (via an OSC 52 escape — works over SSH/remote too).
2. Split the terminal — Ctrl+Shift+5 (Windows/Linux) or Cmd+\ (macOS).
3. Paste into the new pane and press Enter — the live sidebar runs there.

Lost the banner once Claude takes the screen? Run ccr sidecar --hint to reprint the steps and re-copy the command.

The split is a one-time setup per VS Code window: an attached sidecar picks each new ccr session up automatically, so relaunching prints a short note instead of the banner. And if you do paste the one-liner into a second pane, the older pane stands down by itself — there is never more than one live sidebar per session. Profiles stay independent: a personal ccr and a work ccr <profile> run side by side, each with its own state dir and its own sidebar.

On Windows this is the default inside VS Code (Windows Terminal otherwise opens a separate window, so the in-editor split is nicer). On Linux/macOS, ccr defaults to tmux (which works inside the VS Code terminal too); set CCR_VSCODE=1 to use the split-terminal flow there instead.

Automating the split keystroke itself would need a VS Code extension (the code CLI has no "run command" verb) — out of scope for the zero-dependency core. The clipboard + --hint reduce it to split-and-paste.

Wiring the statusline into Claude Code

In ~/.claude/settings.json:

{ "statusLine": { "type": "command", "command": "ccr statusline" } }

(Install the binary with npm i -g claude-code-runrate rather than using npx here — Claude Code calls the status line frequently, and a resolved binary avoids per-tick latency.)

Development

This project is built BDD-first: the Gherkin in features/ is the source of truth, executed by a hand-rolled zero-dependency harness on top of Node's built-in test runner — a single-file Gherkin parser + runner that supports the practical core of the grammar and rejects everything else loudly rather than mis-parsing it. The harness is available standalone as gherkin-node-test on npm (that repo is the canonical source; test/gherkin.js is a vendored copy). See docs/GHERKIN.md for the grammar, the deliberate limits, and the API.

npm test            # node --test — harness self-tests + feature scenarios
npm run typecheck   # tsc --noEmit over @ts-check'd JS (needs: npm i first)

• No runtime dependencies, ever — it's what lets npx claude-code-runrate install instantly on every OS, including native Windows.
• Dev-only tooling (typescript, @types/node) exists solely for npm run typecheck; it is never installed for consumers, so the runtime promise is untouched.
• All source uses // @ts-check + JSDoc — type-checked, but shipped as plain JS that runs straight from source (no build step, no dist/).

Status & support

ccr is a solo side project. I built it to answer one question for myself — am I about to hit my Claude Code limit, and is clearing context worth it? — and I'm sharing it in case it's useful to you too. I use it daily, but please treat it as best-effort and as-is:

• Issues — I read them all. Triage is usually within a week or two, faster for anything that breaks the core (the economy/statusline math, or a platform that won't run). A minimal repro and your OS + Node version help a lot.
• Fixes — core bugs get priority; nice-to-haves may sit for a while.
• Pull requests — welcome, especially small, focused ones with a test. I review on the same best-effort cadence. For anything large, please open an issue first so we don't both sink time into something I'd want shaped differently.
• No SLA and no roadmap promises. Things land when I have the itch or a good PR shows up. If ccr stops being maintained I'll say so plainly at the top of this file.

If it saved you some token math, that's the whole goal. 🎸

License

MIT © 2026 Bing Ho