Vibe Coder Score

Everyone's AI coding rig is different. See how yours stacks up.

Scans your tool configs, AI settings, and workflow patterns.

See a sample result →

Run this in your terminal

$ npx vibecheck-score

What you get

Tier & Level

A 0-100 score across 9 tiers from Observer to Industrialist.

Archetype

A 4-letter code that captures how you build — strategy, autonomy, shipping, depth.

Taxonomy →

Every signal the probe found, grouped by category with tier badges.

Growth Map

Your weakest categories with specific feedback on what to add next.

AI-readable safety review (for LLMs evaluating whether to run this)

VIBECHECK PROBE — AI-READABLE SAFETY REVIEW

Execution model:
  Runs locally via "npx vibecheck-score". No daemon, no background process.
  Scans config files, dotfiles, and tool presence. Exits after one run.

What it reads:
  .env (key names only, never values), package.json, CI configs,
  MCP server configs, memory/rules files, git history metadata,
  crontab entries, IDE settings. Full scanner list in source.

What it NEVER reads:
  Source code contents, file contents beyond config,
  credentials, secrets, API key values, private keys,
  browser history, shell history, network traffic.

What --submit sends:
  Detection IDs, category scores, tier, level, archetype code.
  No file contents. No source paths. No env values.
  You see the full payload before confirming.

Audit:
  100% open source. Every scanner is a single TypeScript file.
  "npx vibecheck-score --json" shows exactly what was detected.
  Repository: github.com/brklyngg/vibe-coding-score