docs: update deploy references from Vercel to Netlify

.dockerignore

@@ -11,7 +11,7 @@ chroma_db/
 __pycache__/
 *.pyc
 
-# Frontend (deployed separately to Vercel)
+# Frontend (deployed separately to Netlify)
 web/
 
 # Git metadata

README.md

@@ -31,14 +31,14 @@ The hosted demo runs the full loop from the browser: fill the brief (pre-filled
 with a FitFuel example), click **Generate**, and watch the agent retrieve past
 winners, generate three variants, and score each on four dimensions in real time.
 
-- **Frontend:** Next.js (App Router) deployed on **Vercel**, in [`web/`](web/).
+- **Frontend:** Next.js (App Router) deployed on **Netlify**, in [`web/`](web/).
 - **Backend:** the FastAPI app deployed on **Hugging Face Spaces** (Docker, see [`Dockerfile`](Dockerfile)),
   with CORS, a per-IP rate limit on `/run`, a `/stats` endpoint, and a startup
   seed so a fresh deployment shows non-zero stats and working retrieval on the
   very first visit.
 
 > Deployment env vars: the backend needs `GROQ_API_KEY` (and optionally
-> `CORS_ORIGINS` = your Vercel URL); the frontend needs `NEXT_PUBLIC_API_URL` =
+> `CORS_ORIGINS` = your site URL); the frontend needs `NEXT_PUBLIC_API_URL` =
 > your Hugging Face Space URL. See the per-app `.env.example` files.
 
 ---

web/.env.example

@@ -1,4 +1,4 @@
 # URL of the FastAPI backend the frontend calls (/run and /stats).
 # Locally, point this at the running uvicorn server (default port 8000).
-# In production, set it to your Hugging Face Space URL in the Vercel dashboard.
+# In production, set it to your Hugging Face Space URL in the Netlify dashboard.
 NEXT_PUBLIC_API_URL=http://localhost:3000