Track Every Token You Save With contextzip gain

How much context are you actually saving? ContextZip tracks it.

$ contextzip gain
Today:     42,891 chars saved (23 commands)
This week: 287,432 chars saved (156 commands)
All time:  1,247,891 chars saved (892 commands)

Every command that runs through ContextZip records its before/after size. contextzip gain shows you the cumulative numbers.

Why This Matters

"60-90% savings" sounds abstract until you see the concrete numbers from your own workflow. After one day of normal development:

• 23 commands ran through the proxy
• 42,891 characters of noise were stripped
• That's roughly 10,000 tokens saved
• At Claude's pricing, that's real money over time

After a week, the numbers compound. A typical development week generates 150-200 CLI commands. At 60-80% noise per command, you're looking at 200K-300K characters saved per week.

Per-Command Visibility

Every command shows its savings inline:

$ npm test
PASS src/utils.test.ts
PASS src/api.test.ts
Tests: 24 passed, 24 total
💾 contextzip: 8,421 → 312 chars (96% saved)

You see the savings in real-time, for every command.

The Graph

$ contextzip gain --graph
Mon ████████████░░░░░░░░ 31,204
Tue ██████████████████░░ 52,891
Wed ████████████████░░░░ 44,123
Thu █████████████░░░░░░░ 38,902
Fri ██████████████████████ 61,203

A daily bar chart of characters saved. Useful for understanding which days generate the most CLI noise (build days, deployment days).

cargo install contextzip
eval "$(contextzip init)"

GitHub: github.com/contextzip/contextzip

---

Part of the ContextZip Daily series. Follow for daily tips on optimizing your AI coding workflow.

Install: npx contextzip | GitHub: jee599/contextzip