← Back to Discover

Vibe Coding Profiler

Web app that reads commit metadata from your repositories and turns it into a shareable persona describing how you work with AI coding tools.

Data analysis

What it does

Vibe Coding Profiler produces a Vibe Coding Profile, a persona for your AI-assisted engineering style based on patterns in your commit history. The site describes it as a playful side project, kept light and approachable for non-technical people, and the version shown is 0.1.0. The site offers a Start with GitHub button as the way in. The tool samples up to 300 commits across a repository's lifetime, using timestamps, sizes and file paths but no code content, and filters out bot activity such as Dependabot and release workflows. It rolls more than 25 metrics into six axes, covering things like planning, guardrails, rhythm and surface area, and maps them to a persona. Examples shown include Spec-Driven Architect, Test-First Validator, Vibe Prototyper and Agent Orchestrator. Each repository gets its own profile, and several repository profiles are combined into a unified one weighted by commit volume. You get share-ready cards plus deeper metrics and evidence. An optional language-model narrative summarizes patterns using metadata only, never code or message content, and only if you opt in.

Capabilities

  • Persona built from commit history signals, not code content
  • Per-repository profiles that roll up into one unified profile
  • 25+ metrics grouped into six axes
  • Share-ready cards and deeper metrics with supporting evidence
  • Bot and automation commits filtered out
  • Optional language-model narrative that uses metadata only

When to use it

  • Seeing how your commit habits changed after adopting AI coding tools
  • Getting a lightweight, shareable summary of your working style
  • Comparing your approach across several repositories