The-One-Env

"One environment to rule them all, one environment to find them, one environment to bring them all and in the darkness bind them."

A unified Python environment to rule all my AI project. Includes all essential dependencies, categorized and described for quick reference and easy setup.

Overview

This environment is designed to provide a reliable foundation for any AI, data, or automation project in my workspace. Use it as a fallback or main environment to ensure all tools are available and compatible.

Environment Management

The-One-Env uses pipenv for environment management because pipenv keeps all virtual environments in a centralized location. This makes it ideal for a unified workspace shared across multiple projects, ensuring consistent dependency management and easy activation from any project folder.

Once uv (a fast Python package manager) supports centralized virtual environment storage, it will be added to the repository. Both uv and pipenv will then be available as options for managing the environment.

Dependencies

Environment

• environs: Environment variable management.
• python-dotenv: Loads environment variables from .env files.

Networking

• httpx: Asynchronous HTTP client.
• requests: Synchronous HTTP requests library.

Data Validation

• pydantic: Data validation and settings management.

Models SDK

• groq: Groq API client for high‑performance inference.
• ollama: Local LLM inference via Ollama.
• openai: OpenAI API client.
• xai-sdk: Explainable AI SDK.
• tiktoken: Tokenizer for OpenAI models.
• anthropic: Anthropic API client.
• google-genai: Google Generative AI SDK.
• deepgram-sdk: Deepgram speech‑to‑text SDK.
• huggingface_hub: Access and manage Hugging Face models and datasets.

Agents SDK

• crewai: CrewAI framework for orchestrating LLM agents.
• openai-agents: OpenAI agents SDK.
• langchain: Unified framework for LLM applications.
• langchain-core: Core utilities for LangChain.
• langchain-chroma: Chroma vector store integration.
• langchain-openai: OpenAI integration for LangChain.
• langchain-ollama: Ollama integration for LangChain.
• langchain-anthropic: Anthropic integration for LangChain.
• langchain-community: Community‑maintained LangChain components.
• langchain-text-splitters: Text splitting utilities.
• langchain-huggingface: Hugging Face integration for LangChain.
• langchain-experimental: Experimental LangChain features.
• langgraph: Graph‑based LLM workflows.
• langgraph-checkpoint-sqlite: SQLite checkpointing for LangGraph.
• langsmith: Experiment tracking for LangChain.

Microsoft autogen (Not to be confused with "Autogen" AKA "AG2")

• autogen-ext: Autogen extensions.
• autogen-core: Autogen core library.
• autogen-agentchat: Autogen agent chat utilities.

UI Framework

• flask: Lightweight web framework.
• gradio: Build machine learning web apps easily.
• fastapi: High‑performance async web framework.
• aiohttp: Asynchronous HTTP client/server.

Web Scraping

• lxml: XML and HTML parsing.
• playwright: Browser automation.
• markdownify: Convert HTML to Markdown.
• beautifulsoup4: BeautifulSoup for parsing.

Document Processing

• pypdf: PDF file reading and manipulation.

Image Processing

• pillow: Image processing library.

Audio Processing

• pyaudio: Audio input/output library.
• simpleaudio: Simple audio playback.

Messaging

• resend: Email sending service.
• sendgrid: Email sending service.

Platform

• modal: Serverless platform.
• wandb: Weights & Biases experiment tracking.
• litellm: Low‑latency LLM inference.

MCP

• mcp: MongoDB Cloud Proxy CLI.

Vector Databases

• chromadb: Chroma vector database.

Lexical Databases

• opensearch-py: OpenSearch client.

Ranking

• rank-bm25: BM25 ranking algorithm.

Transformers

• transformers: Hugging Face Transformers.
• sentence-transformers: Sentence‑embedding models.

PyTorch Computing

• torch: PyTorch core library.
• torchvision: PyTorch vision utilities.

Dev Packages

• isort: Python import sorting tool.
• mypy: Static type checker for Python.
• mccabe: Complexity checker for Python code.
• pyflakes: Checks Python source files for errors.
• pydocstyle: Checks compliance with Python docstring conventions.
• pycodestyle: Checks Python code style.
• pytest: Python testing framework.
• coverage: Measures code coverage for tests.
• ipykernel: Jupyter kernel for Python.
• ipywidgets: Interactive widgets for Jupyter notebooks.
• setuptools: Python package build and distribution.
• pyreadline3: Interactive REPL support.

Using the activation scripts in other projects

To use The-One-Env from any other project, follow these steps:

• Step 1: Generate activation scripts in The-One-Env folder

  • For bash (Linux/Mac):
    • Run: scripts/create-activation-script.sh
  • For PowerShell (Windows):
    • Run: scripts/create-activation-script.ps1
  • This will create activate-the-one-env.sh and/or activate-the-one-env.ps1 in The-One-Env folder.

• Step 2: Use the generated activation scripts in any project

  • Copy the generated activation script(s) to your target project folder.
  • To activate The-One-Env from another project:
    • Bash (Linux/Mac):
      • Run: source ./activate-the-one-env.sh
    • PowerShell (Windows):
      • Run: ./activate-the-one-env.ps1

This will activate The-One-Env environment locally, allowing you to use its dependencies from any location.

Ensure the path inside the script matches your actual .virtualenvs path.