Orano (Ontology-driven Reasoning with AI for Natural Operations) Bot Project
STILL IN IDEATION PHASE
A library to benchmark various Large Language Models (LLMs) by receiving a prompt from the user, converting it into step-by-step actions, and interacting with a computer environment.
Explore the Docs »Report Bug · Request Feature
Table of Contents
-
Vision
- Captures screenshots and detects objects (e.g., buttons, windows, text fields).
- Provides metadata about screen elements.
-
Mind
- Manages prompt handling and interaction with one or more LLMs.
- Breaks user requests into smaller, sequenced commands.
-
OranoBot
- Orchestrates the workflow:
- Accepts user prompts.
- Calls Mind to generate actions.
- Uses Vision to provide environment context (screenshots).
- Executes each step on the computer.
- Orchestrates the workflow:
-
UI (Optional)
- Helper module for user-interface operations, hooking into the OS or external libraries for automation.
- Just Started: The code structure is set up, but actual implementations are minimal or placeholders.
- MLOps: Will be integrated later for continuous model training/deployment, if needed.
- Docker & AWS: Planned, but not implemented yet.
- We maintain a development log in the
devlog/folder.- Each date has its own
.mdfile (e.g.,devlog_2024_12_30.md). - The latest development notes and images are there.
- Each date has its own
- We may occasionally highlight the most recent devlog entries here in the main README.
- You can also follow the project on Medium.
orano/
├── .gitignore
├── LICENSE
├── README.md
├── devlog/
│ └── imgs/
├── docs/
│ └── architecture.md
├── requirements.txt
├── src/
│ └── orano/
│ ├── __init__.py
│ ├── mind.py
│ ├── vision.py
│ ├── bot.py
│ └── ui.py
└── tests/
├── __init__.py
├── test_mind.py
├── test_vision.py
├── test_bot.py
└── test_ui.py
Notes:
src/orano/houses your main code modules.tests/contains unit tests for each module.devlog/stores the development journal and images.
git clone https://github.com/Gabriel382/orano.git
cd oranopip install -r requirements.txt(If you prefer Poetry or Conda, adjust accordingly.)
pytest tests(Or your test runner of choice.)
Check out the devlog/ for detailed progress notes and any relevant screenshots or diagrams.
- Check out our LICENSE to see how this project is distributed.
- Feel free to open issues or pull requests if you have ideas or bug fixes.
- As we grow, we’ll add a
CONTRIBUTING.mdfor guidelines and best practices.
This project is licensed under the MIT License. Please see the license file for full details.
Gabriel Henrique Alencar Medeiros
Happy Building!
Stay tuned for updates, and check our devlog for day-to-day progress.