Skip to content

LuisESV/cocode

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CoCoDe: Code-focused AI Assistant

CoCoDe is a terminal-based AI coding assistant powered by GPT-4.1. It provides a Claude Code-like experience for developers, with the ability to understand codebases, edit files, interact with Git, and more.

Features

  • Code Execution: Run code in various languages directly in the terminal
  • File Editing: Edit files with precise control and preview changes before applying
  • Git Integration: Stage, commit, and push changes with AI-generated commit messages
  • Codebase Understanding: Parse and analyze your codebase for better context
  • Persistent Context: Save and resume conversations by task or file
  • Beautiful Terminal UI: Navigate your codebase and interact with the assistant in a polished TUI

Installation

# Clone the repository
git clone https://github.com/yourusername/cocode.git
cd cocode

# Install the package
pip install -e .

# Or install dependencies directly
pip install -r cocode/requirements.txt

Usage

# Configure CoCoDe with your API key
cocode setup

# Run CoCoDe in the current directory
cocode run

# Run with a specific model
cocode run --model gpt-4o

Requirements

  • Python 3.9 or higher
  • OpenAI API key for GPT-4.1
  • Git (for Git integration features)

Architecture

CoCoDe is built on several key components:

  • Model Connector: Interface with GPT-4.1 API
  • Interpreter Wrapper: Adapted from Open Interpreter for code execution
  • Git Agent: Integration with Git for version control
  • File Indexer: Parse and index the codebase
  • Memory Manager: Manage conversation history and context
  • TUI Interface: Terminal UI built with Textual

Credits

CoCoDe is inspired by Claude Code and leverages the Open Interpreter project for code execution capabilities.

About

Codex before Codex

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages