VibeDrift
Stop AI agents from making codebases worse by detecting and preventing architectural and stylistic 'drift'.
At a Glance
- AI-augmented development teams
- Individual developers using Cursor/Claude Code
- Enterprise engineering teams
AI Tools by VibeDrift
(1)
VibeDrift
CLI Tool for Codebase Drift Detection
Discussions
No discussions yet
Be the first to start a discussion about VibeDrift
Latest News
VibeDrift featured on Hacker News Show HN.
VibeDrift goes open source with initial public release.
VibeDrift MCP server launched for Claude Code and Cursor.
Introduction of the Vibe Drift Score to measure AI technical debt.
Products & Services
A command-line tool to scan codebases for consistency and calculate a Vibe Drift Score (0-100).
A Model Context Protocol (MCP) server that allows AI agents like Claude Code and Cursor to check codebase conventions mid-task.
AI-powered analysis that catch duplicates and intent mismatches with copy-ready fixes.
Market Position
Unlike linters that focus on syntax or static analysis tools that focus on bugs, VibeDrift focuses on 'drift'—the structural and behavioral inconsistency that occurs when multiple stateless AI sessions contribute to a project.
Leadership
Founders
Sami Ahmad Khan
Software Engineer at Meta. Developer of Vibelang and various open-source projects. Based in San Francisco, CA.
Anishek Kamal
Senior Cloud Solution Architect at Microsoft. Over 10 years of experience in Data and AI solutions. Based in San Diego, CA.
Executive Team
Sami Ahmad Khan
Co-Founder
Software Engineer at Meta.
Anishek Kamal
Co-Founder
Senior Cloud Solution Architect at Microsoft.
Founding Story
Built by founders who noticed that the more they used AI to code, the more inconsistent and difficult to maintain their codebases became. They created VibeDrift to give agents the 'shared memory' they lack.
Business Model
Revenue Model
Freemium SaaS offering subscription-based deep scans and credit top-ups for advanced AI analysis.
Pricing Tiers
Unlimited local scans, unlimited MCP tools, 1 deep scan per month.
12 deep scans per month. Credits never expire.
5 additional deep scans.
Target Markets
- AI-augmented development teams
- Individual developers using Cursor/Claude Code
- Enterprise engineering teams
- Preventing technical debt in AI-generated codebases
- Maintaining architectural consistency across large projects
- In-loop convention checking for AI coding agents