EnvAgent

AI-powered tool that automatically generates Conda environments for Python projects

EnvAgent analyzes your Python codebase and creates production-ready Conda environments automatically. It uses GPT-4 for intelligent dependency detection, handles version conflicts, and self-heals installation errors.

---

Quick Start

# Install dependencies
pip install -r requirements.txt

# Configure OpenAI API key
echo "OPENAI_API_KEY=your-api-key-here" > .env

# Generate environment for your project
python main.py /path/to/your/project

# Activate and use
conda activate your_project_name

---

Features

• Smart Dependency Detection - Combines AST parsing with config file analysis for accurate results
• Active Hardware Detection - Detects NVIDIA GPU vs Apple Silicon and optimizes accordingly
• Smart Warnings - Alerts you if your hardware (e.g., Mac) doesn't match project requirements (e.g., CUDA)
• Monorepo Support - Automatically identifies the correct project root in complex repositories
• Self-Healing - Fixes installation errors automatically (up to 8 retries with GPT-4)
• Platform-Aware - Excludes incompatible packages (e.g., CUDA on macOS)
• Token-Efficient - Handles large codebases (1000+ files) without hitting LLM token limits
• Production-Ready - Generates validated environment.yml with loose version constraints

---

Installation

Prerequisites

• Python 3.8+
• Conda (Download Miniconda)
• OpenAI API Key (Get one here)

Setup

git clone https://github.com/yourusername/EnvAgent.git
cd EnvAgent
pip install -r requirements.txt
cp .env.example .env
# Edit .env and add your OPENAI_API_KEY

---

Usage

Basic Commands

# Analyze current directory
python main.py .

# Analyze specific project
python main.py /path/to/project

# Custom environment name
python main.py /path/to/project -n my_env

# Specify Python version
python main.py /path/to/project --python-version 3.10

# Generate YAML only (don't create environment)
python main.py /path/to/project --no-create

Example Workflow

# Analyze a YOLOv5 project
python main.py ~/projects/yolov5

# Output:
# 🔍 Checking system requirements...
# 📋 Analyzing project structure...
# 📁 Filtering source files...
# 🔬 Scanning dependencies...
# 🔨 Generating environment.yml...
# 🚀 Creating conda environment...
# ✅ SUCCESS! Environment 'yolov5' created.

# Use the environment
conda activate yolov5
python train.py --data coco.yaml

---

How It Works

EnvAgent uses a 6-step pipeline to generate environments:

1. System Check - Validates Python, Conda, and Active GPU Hardware (NVIDIA/Apple Silicon)
2. Decision Agent - Detects monorepos and chooses analysis strategy
3. File Filter - Selects relevant Python files and config files
4. Code Scanner - Extracts dependencies using AST parsing + config analysis
5. Environment Builder - Generates OS-aware environment.yml with GPT-4
6. Auto-Fix Loop - Creates environment and fixes errors automatically

See FLOW_ARCHITECTURE.md for detailed technical documentation.

---

Documentation

• FLOW_ARCHITECTURE.md - Detailed execution flow with step-by-step diagrams
• ARCHITECTURE.md - System architecture and technical details
• CONTRIBUTING.md - Contribution guidelines