Gemini Code Flow 🚀

AI-powered development orchestration for Gemini CLI - adapted from Claude Code Flow by ruvnet

🌟 Overview

Gemini Code Flow brings the powerful SPARC methodology and multi-agent orchestration capabilities to Google's Gemini CLI platform. This tool enables developers to leverage multiple AI agents working in parallel to write, test, and optimize code systematically.

🙏 Attribution

This project is adapted from the excellent Claude Code Flow by ruvnet. Full credit goes to the original author for the innovative SPARC methodology and multi-agent orchestration concepts.

✨ Features

• Multi-Agent Orchestration: Run up to 10 Gemini agents in parallel
• 17 SPARC Development Modes: Specialized agents for every development need
• Persistent Memory: Share knowledge across agents and sessions
• Multimodal Support: Leverage Gemini's ability to process images, PDFs, and sketches
• Large Context Window: Utilize Gemini's 1M token context for complex codebases
• Google Search Integration: Ground responses with real-time information

🛠️ SPARC Development Modes

1. 🏗️ Architect - System design and architecture
2. 🧠 Auto-Coder - Clean, modular implementation
3. 🧪 Tester (TDD) - Test-driven development
4. 🪲 Debugger - Troubleshooting and bug fixes
5. 🛡️ Security Reviewer - Security audits
6. 📚 Documentation Writer - Comprehensive documentation
7. 🔗 System Integrator - Component integration
8. 📈 Deployment Monitor - Post-launch monitoring
9. 🧹 Optimizer - Performance optimization
10. ❓ Ask - Task formulation guide
11. 🚀 DevOps - Deployment and infrastructure
12. 📘 SPARC Tutorial - Interactive learning
13. 🔐 Database Admin - Database management
14. 📋 Specification Writer - Requirements and pseudocode
15. ♾️ MCP Integration - External service connections
16. ⚡️ SPARC Orchestrator - Complex workflows
17. 🎨 Designer - UI/UX with multimodal capabilities

📦 Installation

Prerequisites

• Node.js 18+
• Gemini CLI installed globally

# Install Gemini CLI first
npm install -g @google/gemini-cli

# Install Gemini Code Flow
npm install -g gemini-code-flow

🔐 Authentication Options

Gemini Code Flow supports multiple authentication methods:

Option 1: Personal Google Account (Recommended for Individual Use)

Free usage with generous limits:

• 60 requests per minute
• 1,000 requests per day
• Access to Gemini 2.5 Pro with 1M token context

# Simply login with your Google account when prompted
gemini-flow init --sparc
# No API key configuration needed!

Option 2: API Key (For Professional/Enterprise Use)

For multiple simultaneous agents or usage-based billing:

# Set your API key from Google AI Studio or Vertex AI
export GEMINI_API_KEY="your-api-key-here"

Current Limitations (Preview)

• Personal Google accounts: 60 requests/minute, 1,000/day
• Some advanced features may require API key authentication
• For the latest usage limits and capabilities, see Google's official announcement

🚀 Quick Start

# Initialize a new project with SPARC
gemini-flow init --sparc

# Start the orchestrator
gemini-flow start

# Run a specific SPARC mode
gemini-flow sparc architect "Design a microservices architecture"

# Launch multi-agent workflow
gemini-flow sparc "Build a REST API with authentication"

💡 Usage Examples

Simple Task

gemini-flow agent "Refactor this function to use async/await"

Complex Multi-Agent Workflow

gemini-flow sparc orchestrator "Create a full-stack application with React frontend and Node.js backend, including tests and documentation"

Multimodal Development

gemini-flow sparc designer "Convert this wireframe sketch to React components" --file ./wireframe.png

🔧 Configuration

Create a .gemini-flow.json in your project root:

{
  "maxAgents": 10,
  "memoryPath": "./gemini-memory.json",
  "authMethod": "google-account",
  "modes": {
    "architect": {
      "temperature": 0.7,
      "maxTokens": 8000
    }
  }
}

Configuration Options:

• authMethod: "google-account" (default) or "api-key"
• apiKey: Only needed if using "api-key" method
• maxAgents: Number of parallel agents (consider rate limits)
• memoryPath: Location for persistent memory storage

🤝 Contributing

We welcome contributions! Please see our Contributing Guide for details.

📄 License

MIT License - see LICENSE file for details.

🙌 Acknowledgments

• Original Claude Code Flow by ruvnet
• Google Gemini team for the excellent CLI tool
• All contributors and users of this project

📚 Documentation

For detailed documentation, visit our Wiki.

🐛 Issues

Found a bug or have a suggestion? Please open an issue.

Built with ❤️ by the community, adapted from Claude Code Flow