AI-Navi Vanilla

🧠 Screen-aware AI companions for games and beyond.

AI-Navi Vanilla is a lightweight framework for integrating AI assistant logic into visual environments using screen-based OCR and LLM prompts. Designed for modders, developers, and futurists, it enables an immersive, prompt-driven experience that overlays game state with intelligent feedback, prophecy-style prompts, or guidance narration.

---

✨ Features

• 🔲 Zone-Based Prompt Triggering
  Define screen areas that map to different in-game events or HUD elements (via prompt_zones.json)

• 🧠 LLM-Powered Interactions
  Use captured screen text + predefined context to trigger LLM prompts (ChatGPT, Claude, etc.)

• 👁 OCR-First Design
  Uses Tesseract or EasyOCR to extract in-game or screen text dynamically

• 🧩 Mod-Friendly Architecture
  Plug AI-Navi into other mods or tools using lightweight JSON templates

• 🛠️ Completely Offline-Compatible (Optional)
  Swap in local models and OCR engines for no-internet use

---

🧰 How It Works

1. Screen Capture
   Continuously (or on hotkey) captures a defined region of the screen

2. Text Recognition (OCR)
   Extracts readable text from screen zones via OCR

3. Prompt Resolution
   Matches extracted text to zone definitions (prompt_zones.json)

4. LLM Prompt Execution
   Sends text + prompt logic to a connected language model

5. Response Handling
   Displays or routes response to voice, overlay, log, or file

---

🚀 Getting Started

1. Clone the repo

   git clone https://github.com/yourname/ai-navi-vanilla
   cd ai-navi-vanilla

2. Install dependencies

   pip install -r requirements.txt

3. Run the assistant

   python src/main.py

4. Customize your zones
   Edit templates/prompt_zones.json to define capture areas and associated prompts.

---

🧪 Example Use Case

You're playing a game with an intricate UI. When you open your inventory, a prompt zone is triggered that sends screen text to GPT, asking:
"Analyze this loadout. What's missing for a balanced mid-game build?"

The LLM replies with specific suggestions-right as you're thinking about your next move.

---

🔭 Roadmap & Vision

• 🎮 Game-specific mod integrations (Skyrim, Factorio, etc.)
• 🗣️ Voice integration (speech-to-text + TTS response)
• 📱 Mobile or AR client support
• 🌐 Prophetic/visionary theme overlays for storytelling
• 🎥 TikTok/Shorts demos to showcase "AI as oracle" gameplay

---

📄 License

This project is licensed under the Polyform Noncommercial License 1.0.0.
This means you are free to use, modify, and share the code for noncommercial purposes.
Commercial use is not permitted without prior permission from the author.

See the LICENSE file for full details.

---

🤝 Contributing

Open to PRs, forks, and discussions. This project is intended for creative remixing and noncommercial experimentation.
Let's build the future of narrative intelligence-together, ethically and openly.