Morality Based Chatbot

Developed during internship at IIT Mandi, currently being used for data collection purpose.

Post-choice projection + RAG

After each choice, players see:

1. Score changes — which moral traits went up/down (deterministic from moralImpact)
2. Ending lean — soft rule-based trajectory label (not a final result)
3. AI insight — Groq Llama + trait-tagged retrieval explains the choice and offers a labeled speculative forecast

Setup: add GROQ_API_KEY (and optional GROQ_MODEL) to backend/.env. See backend/README.md. Without a key, the UI still works using fallback text.

---

---

---

---

---