google-gemini/gemini-cli

This system is a comprehensive command-line interface and agent framework designed for AI-assisted software development. It provides a modular architecture for managing extensions, skills, and MCP servers, while offering a rich interactive UI for chat, code review, and browser automation. The core engine handles model routing, policy enforcement, and secure sandboxing across Linux, macOS, and Windows, enabling safe execution of tools and subagents. Additionally, it includes infrastructure for automated issue triage, performance monitoring, and SDK integration for embedding agent capabilities into other applications.

The project underwent a comprehensive architectural overhaul, introducing a modular, event-driven scheduler and a policy engine to manage tool execution and security. This period established a robust foundation for the CLI, featuring a new slash command system, extensible extension and skill management, and deep integration with VS Code and A2A servers. The work focused on improving developer experience, security, and the overall stability of the Gemini CLI.

94 changes

This period focused on expanding the agent's functional scope through a new skills infrastructure, pluggable authentication, and a comprehensive CLI SDK. Simultaneously, the codebase introduced robust, platform-specific sandboxing for Linux, macOS, and Windows, alongside a new browser automation agent. The work also established automated workflows for repository maintenance and issue triage.

41 changes

google-gemini/gemini-cli was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point the surveyor at a repository you know and see whether you agree with it.