franciskwarteng677/minewatch-ghana

A public-health awareness and educational mining-risk visualization platform for Ghanaian mining communities. # MineWatch Ghana Live Demo: minewatch-ghana.vercel.app MineWatch Ghana is a static educational MVP for public-health awareness and mining-risk visualization in Ghana. It is built for mining workers, mining families, students, health educators, NGOs, mining-safety organizations, and community leaders. ## Short Description MineWatch Ghana explains occupational health risks in simple language and visualizes educational exposure-risk categories for selected mining areas. It focuses on dust inhalation, silica exposure, mercury exposure, chemical exposure, respiratory warning signs, pneumonia risk, pneumoconiosis risk, chronic cough, breathlessness, chest pain, and delayed symptoms. This project is for health education and early awareness. It does not diagnose disease, treat illness, cure illness, confirm disease, predict confirmed disease, replace medical care, or provide official medical surveillance. ## Ask MineWatch AI Ask MineWatch AI is an educational chat assistant for mining-health questions. It works offline by default using a rule-based response engine in `script.js`, so it still works with Live Server and static deployment without any API key or backend. For deployments on Vercel, the assistant can optionally use GroqCloud when `GROQ_API_KEY` is configured. If the API key is missing, the request fails, or the Groq request is unavailable, the assistant automatically falls back to the existing offline rule-based response. The assistant is not a doctor. It does not diagnose, treat, cure, confirm disease, predict confirmed disease, or replace care from qualified medical professionals. For severe or worsening symptoms, it should advise users to seek medical attention as soon as possible. Current design: - Responses are generated locally in `script.js` when offline mode is used. - Optional Groq-backed responses are served through the Vercel serverless route in `api/ask.js`. - The assistant uses keyword and phrase matching for topics such as dust, s …