Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Improve RAG: query expansion, confidence scoring, graceful 3-tier fallback, better prompting

Fix: use OpenAI client with HF router URL (no provider param needed, works with any version)"

Switch to HF Inference API: Qwen3-32B via InferenceClient (same architecture, faster)"

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py

Update engine.py