AIENG

A terminal-based, model-agnostic AI coding assistant, inspired by GPT Codex

Installation

Prerequisites

• Python 3.12 or higher
• uv package manager

Install uv

If you don't have uv installed, get it from uv installation guide

Build from Source

1. Clone the repository:

   git clone https://github.com/markoelez/aieng.git
   cd aieng

2. Install dependencies and build:

   uv sync

3. Install in development mode:

   uv pip install -e .

4. Set up environment variables:

   Create a .env file in the project root:

   cp .env.example .env

   Edit .env and add your API key:

   OPENAI_API_KEY=your_api_key_here
   

5. Run AIENG:

   aieng

Configuration

Switching Models

AIENG is model agnostic. To switch models:

1. Initialize configuration file (if not already created):

   aieng
   # Type /init to create aieng.toml

2. Edit aieng.toml in your project directory:

   # Model Configuration
   model = "gpt-5.1-codex"  # Default GPT Codex model
   api_base_url = "https://api.openai.com/v1"  # OpenAI GPT Codex endpoint
   
   # Examples for different providers:
   # Anthropic: api_base_url = "https://api.anthropic.com/v1"
   # Local: api_base_url = "http://localhost:11434/v1"

3. Update API key in .env:

   OPENAI_API_KEY=your_provider_api_key_here

TODOs

• Improve agent narration between tasks
• Reduce redundant actions

Development Workflow

1. Install dev dependencies

   uv sync --dev

2. Install git hooks (Ruff formatting + Ty type checking)

   uv run pre-commit install

3. Run checks locally

   uv run ruff check .
   uv run ruff format --check .
   uv run ty check