Back to Projects
Developer Tooling

AI Code Review Agent

RAG-powered code review agent that provides context-aware suggestions and automates PR reviews via GitHub Actions.

Tech Stack

PythonRAGLangChainGitHub ActionsFastAPIVector DBOpenAI API

Overview

The AI Code Review Agent uses a RAG architecture to analyze code changes in the context of the broader codebase. When a pull request is opened, a GitHub Actions workflow triggers the agent, which retrieves relevant code context from a vector index of the repository, then generates detailed, actionable review suggestions. The agent understands patterns, potential bugs, style inconsistencies, and security issues specific to the project's own conventions.

Key Highlights

  • RAG architecture indexes the full repository for project-specific, context-aware reviews

  • Automated PR review triggered via GitHub Actions on every pull request

  • Detects bugs, anti-patterns, and security issues with project-convention awareness

  • Configurable review depth — quick lint pass or deep semantic analysis

  • Review comments posted directly to the PR via GitHub API