fix: bind uvicorn to Railway's dynamic $PORT instead of hardcoded 8000

Railway assigns a dynamic $PORT and runs its healthcheck against it. The
Dockerfile hardcoded --port 8000, so every deploy after healthcheckPath was
added to railway.json failed with "service unavailable" on /health. #316 and
#318 have both been stuck for days; prod only survived on the pre-healthcheck
#293 image until that replica was knocked out. Bind ${PORT:-8000} (fallback for
local/compose) and make the internal healthcheck read the same port.

Author: DevanshuNEU

backend/Dockerfile

@@ -22,13 +22,15 @@ COPY . .
 # Create repos directory for cloning
 RUN mkdir -p repos
 
-# Expose port
+# Expose the default port (informational; Railway injects its own $PORT at runtime)
 EXPOSE 8000
 
-# Health check
+# Health check hits whatever port the app bound to, so it tracks $PORT not a hardcoded 8000
 HEALTHCHECK --interval=30s --timeout=3s --start-period=5s --retries=3 \
-    CMD python -c "import urllib.request; urllib.request.urlopen('http://localhost:8000/health').read()" || exit 1
+    CMD python -c "import os,urllib.request; p=os.environ.get('PORT','8000'); urllib.request.urlopen('http://localhost:'+p+'/health').read()" || exit 1
 
-# Run with uvicorn (proxy-headers enabled for rate limiting behind reverse proxy)
-# Set FORWARDED_ALLOW_IPS env var in production to your proxy's IP range
-CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000", "--proxy-headers"]
+# Bind Railway's dynamic $PORT (fallback 8000 for local/compose). The platform healthcheck
+# queries $PORT, so hardcoding 8000 here makes the deploy fail "service unavailable".
+# exec keeps uvicorn as PID 1 so it gets SIGTERM for graceful shutdown.
+# proxy-headers for rate limiting behind the reverse proxy (set FORWARDED_ALLOW_IPS in prod).
+CMD exec uvicorn main:app --host 0.0.0.0 --port ${PORT:-8000} --proxy-headers