Skill: hybrid-reasoning (Hybrid AI)

Purpose

Dual-path architecture: a DeterministicEngine handles well-defined logic
(reliable, fast, auditable) and an LLMReasoner handles ambiguous/fuzzy cases.
The HybridSystem routes each query to the right path.

Components

Quick Start

from agents.hybrid_reasoning import DeterministicEngine, LLMReasoner, HybridSystem, Rule

system = HybridSystem()
system.set_deterministic_rules([
    Rule("greet", lambda i: i.get("text") == "hello", lambda i: "Hi there!", priority=10),
])
result = system.query({"text": "hello"})
print(result.answer)         # "Hi there!"
print(result.mode.value)     # "deterministic"

Testing

pytest tests/test_hybrid_reasoning.py -v

Branch

fix/quick-note-237-hybrid-ai