Skip to content

Latest commit

 

History

History
 
 

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 

README.md

CLI Coding Agents

🖥️ Memory-powered CLI development with Claude Code and Gemini CLI

Overview

This configuration demonstrates Cipher as a memory layer for Claude Code and Gemini CLI. Unlike traditional CLI tools that lose context between sessions, this setup provides persistent memory that grows with your development workflow.

Key Benefits:

  • Persistent memory across CLI sessions
  • Project-aware assistance that remembers your codebase
  • Cross-session learning that builds on previous interactions

Prerequisites

Required API Keys:

  1. Anthropic API Key - Get from console.anthropic.com
  2. OpenAI API Key - Get from platform.openai.com (required for embeddings)
  3. Google AI API Key (optional) - Get from Google AI Studio

Setup

1. Environment Setup

Set your API keys:

export ANTHROPIC_API_KEY=your_anthropic_api_key
export OPENAI_API_KEY=your_openai_api_key
export GOOGLE_AI_API_KEY=your_google_ai_api_key  # optional

2. Claude Code Configuration

Create or edit .mcp.json in your project root:

{
  "mcpServers": {
    "cipher": {
      "command": "cipher",
      "args": ["--mode", "mcp", "--agent", "./cipher.yml"],
      "env": {
        "ANTHROPIC_API_KEY": "your_anthropic_api_key",
        "OPENAI_API_KEY": "your_openai_api_key"
      }
    }
  }
}

3. Gemini CLI Configuration

Add to your Gemini CLI settings.json:

{
  "mcpServers": {
    "cipher": {
      "command": "cipher",
      "args": ["--mode", "mcp", "--agent", "./cipher.yml"],
      "env": {
        "ANTHROPIC_API_KEY": "your_anthropic_api_key",
        "OPENAI_API_KEY": "your_openai_api_key"
      }
    }
  }
}

4. Test the Setup

Claude Code:

claude
> Analyze my project structure and remember the patterns

Gemini CLI:

gemini "Use cipher to analyze my codebase"

Usage Examples

Project Analysis:

claude> Analyze my project structure and remember the patterns

Cross-Session Learning:

# Session 1
claude> Help me debug this React performance issue

# Session 2 (later)
claude> Apply the optimization techniques we discussed yesterday

Code Review:

claude> Review my authentication changes using the security patterns you've learned

This setup provides persistent memory for your CLI development workflow, making your coding agents smarter with every interaction.