Semantically compressed RAM-based language exchange
Featuring rambleMAXX: the AI enhanced terminal chat interface from the 1980's, now with Magic Scroll technology!
- Multi-model chat client - Communicate with both local and remote LLMs seamlessly
- Context-aware chat - Smart retrieval of relevant conversation history
- SQLite-powered database - Self-contained storage with vector search capabilities
- Terminal-based UI - Clean, efficient interface for maximum productivity
From source:
# Clone the repository
git clone https://github.com/yourusername/scramble.git
cd scramble
# Install dependencies and the package in development mode
pip install -e .Start the interactive CLI:
./tools/start_ramble.shOr take it to the MAXX with rambleMAXX:
maxx- Python 3.8+
- sentence-transformers
- sqlite-vec (automatically installed with package)
- ollama (optional, for local models)
scRAMble uses a SQLite database with vector search capabilities to store and retrieve chat conversations. This approach offers several advantages:
- Self-contained - Everything is stored in a single file (~/.scramble/magicscroll.db)
- Fast retrieval - Vector similarity search for finding relevant context
- No external dependencies - No need to run separate database servers
See DEVELOPMENT.md for development guidelines.
For SQLite vector search setup, see docs/SQLITE_SETUP.md.
MIT