Open Code Review

Version updated for https://github.com/raye-deng/open-code-review to version v2.1.0.

  • This action is used across all versions by 0 repositories.

Action Type

This is a Node action using Node version 20.

Go to the GitHub Marketplace to find the latest changes.

Action Summary

Open Code Review is an open-source CI/CD quality gate designed specifically for analyzing AI-generated code across six programming languages. It automates the detection of issues such as hallucinated imports, stale APIs, over-engineering, context window artifacts, and security anti-patterns, which traditional static analysis tools often miss. The action provides fast, scalable, and self-hostable code reviews with multi-stage pipelines that include structural analysis, embedding-based checks, and deep remote LLM analysis.

Release notes

What’s New

L3 Deep Scan with Remote LLM

  • Full L3 SLA pipeline: Structural → Embedding → Remote LLM
  • 8 LLM providers: OpenAI, GLM, ZAI, DeepSeek, Together AI, Fireworks, Anthropic, OpenAI-compatible
  • CLI: ocr scan . --sla L3 --provider glm --model pony-alpha-2 --api-key KEY

Universal Provider Adapter

  • openai-compatible adapter for any OpenAI-compatible service
  • Preset providers with auto-filled baseUrl
  • Config file support (ocr.config.json)

Multi-Language AI Detectors

  • Go, Java, Kotlin, Python language-specific detectors
  • Hallucinated imports, over-engineering, security anti-patterns
  • Phantom package databases for each language

ocr heal Command

  • AI-powered auto-fix: scan → fix → re-scan
  • --dry-run mode for preview
  • IDE rules generation (Cursor, Copilot, Augment)
  • Heal report with before/after comparison

Bug Fixes

  • CLI –version now shows correct version
  • SARIF version number fixed