Open Code Review – Detect AI-specific code defects missed by traditional linters

Summary: Open Code Review is a free, open-source CLI tool that identifies five AI-specific code defects undetected by linters like ESLint and SonarQube. It supports multiple languages and integrates with GitHub Actions, using layered analysis including AST, embedding similarity, and LLM deep scans to find issues such as phantom packages and context breaks.

What it does

It analyzes code through structural AST checks, semantic embedding similarity, and LLM-based deep scans to detect hallucinated imports, outdated APIs, hidden anti-patterns, and logic inconsistencies across files. The tool supports TypeScript, Python, Java, Go, Kotlin, and JavaScript, producing SARIF output for code scanning.

Who it's for

Developers using AI coding assistants who need to catch subtle defects that traditional linters overlook in multi-language projects.

Why it matters

It addresses AI-generated code errors that appear correct but contain hidden defects, improving code reliability beyond standard linting tools.