Vibe Coding Detector
Assess whether a feature idea, prompt, website, or GitHub repository feels like rapid vibe coding instead of deliberate product engineering. Paste text, scan any website, or point at a GitHub repo to score vague language, missing safeguards, weak validation, and hacky shortcuts, then get a clear verdict and actionable review notes.
Paste a URL or GitHub repository and get a probabilistic Vibe Coding Score from 0–100, backed by observable evidence — AI coding tool signatures, generated code patterns, CSS density, content signals, SEO and accessibility checks.
Advertisement
Common Use Cases:
- •Reviewing product briefs before implementation
- •Auditing low-quality AI-generated requests
- •Checking whether a developer note feels rushed or vague
- •Scanning websites for AI-generated template patterns
- •Reviewing a GitHub repo before contributing or buying
- •Spotting missing production safeguards in code ideas
Features
- Risk score from 0 to 100 based on quality signals
- Scan any website URL for vibe-coding patterns
- Scan GitHub repos: checks tests, CI, README, and source code
- Flags vague, hacky, and shortcut-heavy language
- Highlights missing tests or validation concerns
- Simple copyable summary report
How to Use
- 1Enter a public website URL or GitHub repository
- 2Watch the 13-stage analysis pipeline
- 3Review the score, evidence, and SEO/accessibility audits
Vibe Coding Detector – Frequently Asked Questions
What is a vibe coding detector?
It is a heuristic tool that scans a website's public source — HTML, scripts, comments and metadata — for patterns statistically associated with AI-assisted development, then reports an evidence-weighted score. It cannot prove that a site was built by AI.
Can you detect if a website was generated with AI?
No detector can prove AI authorship from public source alone. This tool reports observed evidence (AI tool signatures, generated-looking code, template content) with per-signal confidence, and explicitly separates evidence from inference.
Does using Next.js or Tailwind mean a site is AI-coded?
No. Modern frameworks and utility CSS are used by millions of human developers and are never treated as AI evidence on their own. A high score requires multiple independent AI-related signals.
Is the scan safe?
Yes. The scanner blocks private, loopback and cloud-metadata addresses (SSRF protection), enforces timeouts, response-size and redirect limits, and never executes downloaded JavaScript.
What does the score actually measure?
The 0–100 Vibe Coding Score is a weighted combination of seven evidence categories: AI-tool signatures (25%), code patterns (20%), generated structure (15%), CSS patterns (10%), JavaScript patterns (10%), content signals (10%) and boilerplate (10%). Confidence is reduced automatically when evidence is weak.
Related Developer Tools
More free online tools you might find useful: