gemini-cli and gemini-flow
These are ecosystem siblings where A is the underlying Gemini CLI framework that B reimplements and extends as an autonomous multi-agent workflow orchestrator, making B a specialized application layer built on A's terminal integration foundation.
About gemini-cli
google-gemini/gemini-cli
An open-source AI agent that brings the power of Gemini directly into your terminal.
Built on the Gemini API with Model Context Protocol (MCP) support for extensible tool integration, it enables code-aware conversations with real-time web grounding and file/shell operations. Features include multimodal input (PDFs, images), conversation checkpointing, and GitHub Actions integration for automated code reviews and issue triage, while supporting multiple authentication paths: OAuth, API key, or Vertex AI for enterprise workflows.
About gemini-flow
clduab11/gemini-flow
rUv's Claude-Flow, translated to the new Gemini CLI; transforming it into an autonomous AI development team.
Implements dual A2A and MCP protocol support for inter-agent communication, orchestrates all 8 Google AI services (Veo3, Imagen4, Lyria, Chirp, etc.) through unified routing with automatic failover, and deploys as a native Gemini CLI extension. Features a 96-agent hierarchical swarm with Byzantine consensus validation, achieving 396,610 SQLite ops/sec and <75ms routing latency for production-grade autonomous development workflows.
Scores updated daily from GitHub, PyPI, and npm data. How scores work