| import numpy as np | |
| import os | |
| import sys | |
| sys.path.insert(0, os.path.expanduser("~/vitalis_devcore")) | |
| from vitalis_ide.math_core.kernel import VitalisKernel | |
| class InferenceEngine: | |
| def __init__(self): | |
| self.kernel = VitalisKernel() | |
| def reason(self, prompt: str) -> str: | |
| tokens = prompt.strip().split() | |
| vec = self.kernel.vectorize_tokens(tokens) | |
| confidence = float(np.mean(np.abs(vec))) | |
| if "scaffold" in prompt.lower(): | |
| return "scaffold" | |
| elif "write" in prompt.lower() or "fix" in prompt.lower(): | |
| return "write" | |
| else: | |
| return f"[INFER] Confidence={confidence:.3f} | Input={prompt[:80]}" | |
| def embed(self, text: str) -> np.ndarray: | |
| return self.kernel.vectorize_tokens(text.strip().split()) | |