Judges Code Review
Version updated for https://github.com/KevinRabun/judges to version v3.3.0.
- This action is used across all versions by 0 repositories.
Action Type
This is a Composite action.
Go to the GitHub Marketplace to find the latest changes.
Action Summary
The Judges Panel GitHub Action provides automated, multi-dimensional code evaluations for AI-generated code, addressing common failure modes such as security vulnerabilities, poor error handling, and compliance issues. It leverages deterministic pattern matching alongside LLM-powered expert analysis to assess 35 specialized domains, generating plain-language findings, prioritized fixes, and actionable reports. Designed for seamless integration into CI pipelines, it simplifies quality assurance through instant evaluations, auto-generated patches, and support for SARIF output for GitHub Code Scanning.
Release notes
What’s Changed
Unified AST Engine
- Consolidated typescript-ast.ts (TypeScript Compiler API) into tree-sitter-ast.ts — all 8 languages now use a single tree-sitter WASM engine
- Removed the TypeScript Compiler API dependency for JS/TS parsing
- TypeScript and JavaScript parsed via tree-sitter-typescript.wasm (superset grammar)
C++ Language Support
- Added C++ as the 8th supported language for AST analysis
- Language aliases: cpp, c++, cc, cxx, c, h, hpp
- Full support: functions, complexity, nesting, parameters, void*/auto weak type detection, #include imports, classes/structs
Bug Fixes
- Rust: Fixed unsafe fn declaration detection (previously only detected unsafe blocks)
- TypeScript: Fixed require() import detection (CommonJS-style imports now counted)
- Test timing: Moved tree-sitter WASM warm-up before test registration to prevent race conditions
Tests
- 711 tests across 139 suites — all passing
- Added 9 new C++ test cases covering parsing, complexity, nesting, parameters, weak types, imports, classes, and evaluator integration
Supported Languages (AST)
TypeScript, JavaScript, Python, Rust, Go, Java, C#, C++