Lucidia Core
AI reasoning engines for specialized domains - physics, mathematics, chemistry, geology, and more.
pip install lucidia-core
What is this?
Lucidia is a collection of specialized AI reasoning agents, each modeled after a domain expert:
| Agent | Domain | Capabilities |
|---|---|---|
| Physicist | Physics | Energy modeling, force calculations, feedback systems |
| Mathematician | Mathematics | Symbolic computation, proofs, numerical analysis |
| Chemist | Chemistry | Molecular analysis, reactions, compound properties |
| Geologist | Geology | Terrain modeling, stratigraphy, resource mapping |
| Analyst | Data Science | Pattern recognition, insights, statistical analysis |
| Architect | Systems | Design blueprints, architecture planning |
| Engineer | Engineering | Structural analysis, calculations, optimization |
| Painter | Visual | Graphics generation, artistic rendering |
| Poet | Creative | Poetry, lyrical composition, narrative |
| Speaker | NLP | Speech synthesis, communication, translation |
Quick Start
CLI Usage
# List available agents
lucidia list
# Run the physicist agent
lucidia run physicist --query "Model energy flow in a thermal system"
# Start the API server
lucidia api --port 8000
API Usage
# Start the server
lucidia-api
# Or with Python
python -m lucidia_core.api
Then query the agents:
# Check health
curl http://localhost:8000/health
# Physicist analysis
curl -X POST http://localhost:8000/physicist/analyze \
-H "Content-Type: application/json" \
-d '{"query": "Calculate the energy required to heat 1kg of water from 20C to 100C"}'
# Mathematician computation
curl -X POST http://localhost:8000/mathematician/compute \
-H "Content-Type: application/json" \
-d '{"query": "Solve x^2 - 5x + 6 = 0"}'
Python Usage
from lucidia_core import get_physicist, get_mathematician
# Load the physicist
PhysicistSeed, load_seed = get_physicist()
seed = load_seed("codex21.yaml")
# Load the mathematician
MathematicianSeed, load_seed = get_mathematician()
API Endpoints
| Endpoint | Method | Description |
|---|---|---|
/health |
GET | Health check and agent list |
/physicist/analyze |
POST | Physics analysis |
/physicist/energy-flow |
POST | Energy flow modeling |
/mathematician/compute |
POST | Mathematical computation |
/mathematician/prove |
POST | Proof assistance |
/chemist/analyze |
POST | Chemical analysis |
/geologist/terrain |
POST | Terrain analysis |
/analyst/insights |
POST | Data insights |
/architect/design |
POST | System design |
Architecture
lucidia-core/
├── lucidia_core/ # Package module
│ ├── api.py # FastAPI endpoints
│ └── cli.py # CLI entry point
├── physicist.py # Physics reasoning engine (867 lines)
├── mathematician.py # Math reasoning engine (760 lines)
├── chemist.py # Chemistry engine (569 lines)
├── geologist.py # Geology engine (654 lines)
├── analyst.py # Data analysis (505 lines)
├── architect.py # System design (392 lines)
├── engineer.py # Engineering (599 lines)
├── painter.py # Visual generation (583 lines)
├── poet.py # Creative text (250 lines)
├── speaker.py # Speech/NLP (302 lines)
├── codex*.yaml # Agent seed configurations
└── quantum_engine/ # Quantum computing extensions
Configuration
Each agent is configured via a YAML "seed" file (e.g., codex21.yaml for Physicist):
system_charter:
agent_name: "Physicist"
generation: "2024"
domain: ["physics", "energy", "dynamics"]
moral_constant: "preserve momentum gently"
core_principle: "ground every run on measured observations"
License
MIT - See LICENSE for details.
Built by BlackRoad OS