Crit – a local review tool for agent plans and code diffs.

Crit addresses a critical gap in the AI-assisted development workflow: effective review of agent-generated output. As agents become more prevalent in generating code and plans, developers need robust tools to inspect, provide feedback, and iterate. Crit's local, browser-based, GitHub-inspired interface provides a familiar and efficient environment for this. The ability to "juggle parallel conversations with agents" and review "before it hits GitHub" directly improves developer productivity and code quality. The optional self-hostable team feedback mechanism further enhances its utility for collaborative environments. This tool is well-positioned within the emerging ecosystem of AI agent tooling, focusing on human-agent collaboration and quality assurance.