Show HN: Argus – AI code review that doesn't grade its own homework (github.com)

Argus has launched a new AI code review platform that emphasizes independent evaluations, ensuring that the AI writing the code does not also assess its quality. This local-first, modular system combines various tools for structural analysis, semantic search, and LLM-powered reviews, allowing developers to identify issues that typical copilots might overlook. Key features include the use of git history intelligence for comprehensive codebase context and a commitment to avoiding vendor lock-in, as it supports OpenAI, Anthropic, and Gemini with seamless provider switching. This innovation is significant for the AI/ML community as it addresses the critical issue of bias in AI-generated code assessments. By separating the coding and review functionalities, Argus fosters reliability and objectivity in the software development lifecycle. Developers can perform detailed code reviews by leveraging hybrid code search capabilities and structural mapping, thereby enhancing productivity and code quality. The platform's easy setup and integration with existing tools improve accessibility for developers looking to implement AI in their workflows without incurring costs or risking dependency on a single provider.