●1 coding with AI today1 developer153 active developer-days2,243 commits shipped with AIlongest live streak: @hattray, 86 days27.6B tokens
vibestreak. Sign in with GitHub
The AI coding streak board

Show your work,
not the bill.

Prove you build with AI, every day, and help employers find you. One live heatmap of the days you coded with Claude Code, Codex, Cursor, Copilot and Gemini and the commits you shipped, across all your GitHub accounts, machines and private repos, ready for your README, LinkedIn or résumé.

@hattraylongest live streak · Claude Code, Codex, Antigravity
86dcurrent streak
153active days
2,243shipped with AI
Why developers use it

Get found

Employers and clients can see that you build with AI, checked against GitHub where it can be, and browse the board by the tools they hire for. Your profile links your GitHub accounts and shows up in search.

Get credit for private work

GitHub's graph misses private repos, work accounts and the agent behind each commit. Your streak card shows all of it, without exposing a line of code.

Build the habit

A streak, levels and an evening nudge keep you shipping with AI every day, as the tools change month to month.

See your own patterns

Every agent in one heatmap: which tools you use, the days you ship, and how many commits your tokens become.

Leaderboard
Last 30 days
All tools
✓ Verified only

Active days: a day with AI coding or a commit shipped with an AI co-author.

#DeveloperLast 12 weeksActive daysStreakShipped with AITokens
1hattray ✓ ◆Month 2 accounts
Claude Code · Codex · Antigravity
30 86d 420 29d 4.2B

How it works

01

Run the CLI

It reads Claude Code, Codex, Cursor, Copilot, Gemini, Antigravity and OpenCode history, and your git log. Prompts and code never leave your machine.

02

Ship with AI

Every commit with an AI co-author counts as shipped, read from your own repos, private ones included, across all your accounts.

03

Keep the streak

A daily sync keeps your heatmap and streak card live on your profile, GitHub README and LinkedIn.

Use it anywhere

WEB

On the web

Your profile, the leaderboard, and a live streak card for your README, LinkedIn and résumé.

TERMINAL

In your terminal

npx vibestreak shows your streak, heatmap and rank; board and today show the leaderboard and who's coding today. No browser needed. Commands

MCP

In your AI assistant

Connect vibestreak to Claude, ChatGPT or Claude Code and ask "what's my streak?" or "who's top this week?". Set it up

Questions

Why would I want this?

To show that you build with AI, consistently, in a way others can check. Your profile and streak card go on your GitHub README, LinkedIn or résumé; employers and clients can find you on the board by the tools you use; and the streak helps you keep the habit. Your GitHub graph alone can't show any of this: it misses private and work repos, and which commits AI helped write.

Can I check it from my terminal or my AI assistant?

Yes. npx vibestreak shows your streak, heatmap and rank in the terminal, and npx vibestreak board the leaderboard. vibestreak is also an MCP server: connect it to Claude, ChatGPT or Claude Code and ask "what's my streak?" or "who's top this week?". How to connect it.

Can employers see my profile?

Yes, if you want them to. Profiles are public by default, link your GitHub accounts, and appear in search; the verified badge shows that GitHub's own public calendars back up your shipped days. Hide your profile in Settings at any time.

What counts as an active day?

A day you used an AI coding agent, or landed a commit with an AI co-author. Streaks run in your own timezone.

Why rank on consistency instead of tokens or cost?

Spend measures how much a model ran, not what you built. Consistency and shipped commits reward the habit that makes AI-assisted work pay off. Tokens are still on the board for comparison.

What leaves my machine?

Daily totals per tool (messages, tool calls, tokens, model names) and, per commit, a keyed hash, its day and the agent that co-wrote it. Never prompts, code, file paths, repo names or commit messages. Details.

I use several GitHub accounts and machines.

Link them all to one profile. Machines add up, and a commit stored on two machines counts once.