COMPASS is an axiomatic reasoning engine and value system based on universal logical and physical principles. The core goal is to create a system that can evaluate input, generate reasoning paths, and provide structured outputs based on fundamental axioms. COMPASS is modular, extensible, and designed for integration with both classical and LLM-driven architectures (e.g., via Metacortex or Ollama backends).
- Axiomatic value assessment based on a defined axiom set
- Modular codebase with clear separation of logic, LLM-interfaces, and evaluation
- FastAPI backend for programmatic interaction
- Ready for containerized deployment (Docker)
- Prepared for integration with vector models, FNet-style word analysis, and superposition routines
core/
evaluator.py # Core logic for evaluating user input and mapping to axioms
llm_interface.py # Handles LLM/cognitive system interactions
...
main.py # FastAPI app entrypoint
Dockerfile.compass # Docker build configuration
requirements.txt # Python dependencies
README.md # This file
-
Clone the repository:
git clone https://github.com/dwpplumb/COMPASS.git cd COMPASS -
Install requirements:
pip install -r requirements.txt
-
Run the FastAPI server:
uvicorn main:app --reload --host 0.0.0.0 --port 8000
Edit and use the provided Dockerfile.compass and docker-compose.yml to build and run COMPASS together with related containers (e.g., frontend or Metacortex).
POST /evaluate/
{
"text": "What is existence?"
}
Response:
{
"result": ...
}
- Axiomatic Consistency: Every response is mapped to a defined axiom or subaxiom.
- Extendability: New axioms, evaluation modules, and LLM backends can be added modularly.
- Transparency: All evaluation paths and outputs are documented and reproducible.
MIT License (recommendation)
Project by David Plumb / COMPASS System Architecture
Project Board: COMPASS GitHub Project