fix(hf): switch to sdk:docker — bypasses Gradio startup event conflicts, runs uvicorn directly

fix(hf): mount Gradio at root path so HF /startup-events health check passes

fix(hf): add uvicorn.run() to block process and keep HF Space container alive

fix(hf): remove demo.launch() and export app for HF uvicorn auto-detection, fix unhashable dict crash

fix(hf): call demo.launch() to keep HF Space container alive with FastAPI routes mounted

fix(hf): use gr.mount_gradio_app to pass HF supervisor health check and maintain persistent container loop

fix(hf): use persistent asyncio uvicorn server runner in app.py

fix(hf): remove gradio import and pin huggingface-hub<0.24.0 to resolve HfFolder ImportError

feat(hf): serve custom React frontend on root and FastAPI API on /api/v1 (hide Gradio UI)

fix(hf): use app:app string import in uvicorn.run for persistent HF server loop

fix(hf): change colorTo to green in README.md YAML header for HF Space validation

chore: track model files with Git LFS for Hugging Face

feat(hf): add Hugging Face Space metadata frontmatter to README.md

feat(hf): unify React Frontend and FastAPI AI Backend in app.py for single HF Space deployment

feat(hf): add app.py Gradio entrypoint and requirements.txt for Hugging Face Spaces backend deployment

fix(backend): restore real trained MobileNetV2 and EfficientNet-B0 neural network prediction pipeline

fix(vercel): add dual path resolution in api/index.py and main.py for serverless entrypoint

fix(backend): guard jose and passlib imports in auth.py to fix serverless FUNCTION_INVOCATION_FAILED on Vercel

fix(backend): make top-level imports in llm_service and rag_service fully guarded for serverless cold start

fix(backend): bypass heavy torch loading in Vercel serverless environment for zero OOM crashes

fix(backend): add numpy tensor guards in disease and soil models for serverless environment

fix(backend): use safe_get_remote_address and lazy model instantiation for serverless Vercel execution

fix(backend): lazy-initialize cache backend to resolve FUNCTION_INVOCATION_FAILED on Vercel

fix(backend): add resilient fallbacks for disease, soil, and image preprocessing in serverless environment

fix(frontend): simplify CropTab state and inputs to prevent tab freeze

fix(frontend): add finite number validation and safe CSS style bounds in CropTab

fix(frontend): make CropTab safe against null/NaN values to prevent tab freeze

fix(vercel): add index.html and SPA rewrites to resolve 404 NOT_FOUND on Vercel

fix(vercel): optimize serverless dependencies and crop model fallback to pass 500MB function size limit

fix(vercel): add .vercelignore and function excludeFiles to prevent node_modules from inflating Python serverless zip

feat(monorepo): configure optimized fullstack Vercel monorepo deployment

fix(vercel): separate frontend build to bypass 500MB serverless bundle limit

trigger(vercel): force redeploy with updated buildCommand

fix(vercel): add npm install to buildCommand in vercel.json

fix(vercel): pin Python 3.11 and update pydantic version to resolve wheel compilation error

feat: configure Vercel Python serverless function handler for FastAPI backend

build: add root vercel.json for frontend deployment

feat: replace legacy frontend, integrate agriculture-ai-frontend with backend, update CORS, and clean project structure

Add trained ML models (crop, disease, soil), tools, setup scripts, and updated .gitignore

first commit