for
Claude Codeand
Codex
The cheat code for AI-native teams.
A coach for every person, and a shared view for the team. Raise each person’s bar, then connect them.
Feed
Updated just nowFull logs added about 12,000 input tokens per debugging session.
Read only the failing test and the relevant error lines.
Keep the useful context. Skip the repeated output.
Define “done” before the first edit.
Add acceptance criteria so your AI can check its own work.Everyone on your team uses AI.As a team, you’re still guessing.
Each person prompts their own way, at their own level. The tokens and the misses stay in their own chat.
Maya (Engineering)
the checkout test is flaky again. use opus and keep rerunning it until it passes
Reran the suite 14 times.
Wrapped the test in a retry, so it passes now.
$ npm test checkout
✕ checkout total · run 1–13
+ retry(3, () => expect(total).toBe(42))
attempt 3 of 3
checkout total · run 14
This used 1.24M tokens.
Over a million tokens on the most expensive model, and the bug is still there, just hidden behind a retry.
From disconnected AI users to
a connected AI-native team.
As each person’s AI fluency goes up, the links between them get stronger: shared rules, shared skills, one way of working.
Today each person uses AI at a different level, from 3 to 6 out of 10, and the links between them are weak. In an AI-native team, everyone is at 9 or 10 and the links are strong: one shared AI workflow, and team productivity rises with it.
Introducing Vibemetric
One simple app. Everyone better at AI, together.
A coach for every teammate.
See your whole team in one place.
Members choose to share their score. You see who’s strong at what, where the team gets stuck, and who can show the others how.
Acme product team
5 members shared a score
Maya
Engineering
MemberOverall score 48/100
Plans big changes before writing code
Uses the strongest model for small edits
Joe
Marketing
MemberOverall score 61/100
Keeps a brand-voice skill for drafts
Rarely checks drafts against real data
Priya
Design
MemberOverall score 39/100
Shares mockups as context
The design system never reaches the AI
Sam
Operations
MemberOverall score 55/100
Reuses saved prompts for weekly reports
Numbers go out without a check
Alex
Sales
MemberOverall score 46/100
Writes clear, short follow-ups
Pastes customer data into chats
- Each person chooses to share. Their scores and one line per area are all that leaves their Mac.
See what worked.
Vibemetric checks whether each tip worked and shares the wins with the team, so one person’s win becomes everyone’s habit.
Compared with the usual answer
A fraction of the cost of AI consulting.
| Compared on | AI consulting | Vibemetric |
|---|---|---|
| Cost | $25,000 | $19per user, per month for Pro; the app is free |
| Time | 60 days | 1 dayhelps from the first score |
| Approach | Top-down | Every personeach person gets better, every day |
| Over time | Goes stale | Keeps uppeople keep up when the tools change |
AI readiness assessments typically run $10,000–$30,000 in 2026 pricing guides; 60 days is a round figure for an engagement. Pro is $19 per user, per month.
Pricing
Free for everyone. Team when you’re ready.
Free
$0forever
For each person on the team.
- AI fluency score across ten skills
- A coaching feed: what happened in your own sessions, and one thing to do next
- Checks whether each fix worked
- Runs on your Mac. Free and open source.
Team
For teams$19per user, per month
For teams going AI native.
- Everything in Free, for every teammate
- Team view: everyone’s shared score in one place
- Team wins: see which tips worked, and spread them
- Fix prompts and deep dives for each skill
- Only scores and one line per area leave each Mac
Questions
Your sessions stay on your Mac.
How does Vibemetric work out where each person stands?
It reads the Claude Code and Codex history already on each person’s Mac and builds a short digest of their sessions: how they plan, give context, check results and fix problems. That person’s own Claude Code or Codex then scores the digest against ten AI skills, and every reason points back to the sessions it came from.
Do our prompts, code or transcripts go to Vibemetric’s servers?
No. Scoring runs through each person’s own Claude Code or Codex, under their own account, the same as any session they run themselves. Vibemetric’s server never receives sessions, transcripts, prompts or code. It gets only each score’s numbers, to keep the score history.
What happens to API keys and other secrets in our sessions?
Before anything is scored, the digest removes the API keys, tokens, webhook URLs and database connection strings it recognizes. The digest stays on that person’s Mac, next to its result, until they delete the result.
Can Vibemetric change our code or settings?
No. It never writes to ~/.claude or ~/.codex and never changes a transcript. The scoring run can only read files: no shell, no edits, no web.
What does the team view show, and who decides?
Each person decides. Nothing reaches the team until they choose a score and share it, and then only that score and one line per area leave their Mac. Never their prompts, transcripts or code. They can take a shared score back at any time, and leaving the team removes everything they shared.
Is the score a performance review?
No. It’s coaching for each person. An AI reviewer writes it, so it can move a point or two between runs, and every reason cites the sessions it came from so anyone can check it.
Can our security team check what it does?
Yes. The app is free and open source, so anyone can read exactly what it reads and what it sends. It’s signed and notarized by Apple.
Every detail is on the privacy page.
Contact
Bringing it to your team? Talk to us.
Questions about setup, privacy or team pricing. We reply by email.
Stop guessing. Go AI native.
Score how each person works, fix one thing at a time, share it with the team, and spread what works.
Free and open source at vibemetric.app · macOS 14+ · Apple Silicon and Intel · 12 MB download