docs: add comprehensive security guidelines and deployment documentation

fix: read Hugging Face mounted secrets from /var/run/secrets/ in Docker CMD

fix: bind VITE_GEMINI_API_KEY from process.env and GEMINI_API_KEY in Docker startup

chore: restrict LLM model list strictly to 100% Free-Tier Flash models

fix: update active model list to gemini-3.6-flash, 3.5-flash, 3.1-flash-lite, flash-latest

fix: add gemini-2.0-flash model with 1.5-flash fallback and SSE streaming support

fix: update Gemini API model to gemini-1.5-flash and improve error messaging

docs: add live demo link

docs: Correct regional scope in README

docs: Update README with V2 features

features- Enhanced Map and what if simlulation and heat pockets.

Fix logo squishing and aspect ratio in AIChat by using scale and flex-shrink-0

Fix horizontal scrolling issue on mobile for the Ask Ritu AI sidebar

Fix weather card cutoff by adding screen bounds detection and scroll

Fix Vite allowedHosts for Hugging Face and YAML frontmatter formatting

Restore Hugging Face YAML frontmatter configuration

Add Dockerfile and ignore video files

Update Vite base path to root for Hugging Face deployment

Remove YAML frontmatter from README

Update README title to accurately reflect Maharashtra scope

Add Docker and Podman instructions to README

Fix sidebar close button z-index and index.html favicon path

Fix AI Chat expanded modal close button to fully close the sidebar chat

Fix absolute paths for map assets to respect Vite base path

Enhance README: Add Sanskrit quote, Python backend info, and Ritu meaning

Enhance README structure inspired by Nyaya Agent

Initial commit of ISRO Climate Twin

Backup before Problem Statement changes