feat: add Cloudflare Worker implementation for edge-to-edge streaming bypass

- Add worker.js: full API on Cloudflare's edge network (1720 lines)
- Add wrangler.toml: deployment configuration for wrangler CLI
- Streaming endpoints now bypass Cloudflare bot detection via edge-to-edge
- Zero cold starts, 100K free requests/day, no npm dependencies
- 15 endpoints: search, trending, info, episodes, watch, genres, random, etc.
- Self-healing fallback: direct → ScraperAPI → FlareSolverr
- Proper code style: box headers, JSDoc, feature markers, inline notes
- Bump version to 2.3.2 across package.json, server.js, apiRoutes.js

Fixes #2

src/routes/apiRoutes.js

@@ -1124,7 +1124,7 @@ const createApiRoutes = (jsonResponse, jsonError, startTime) => {
       success: true,
       results: {
         status: "healthy",
-        version: "2.3.0",
+        version: "2.3.2",
         uptime: `${hours}h ${minutes}m ${seconds}s`,
         uptimeSeconds: uptime,
         timestamp: new Date().toISOString(),
@@ -1211,7 +1211,7 @@ const createApiRoutes = (jsonResponse, jsonError, startTime) => {
       info: {
         title: "MiruroAPI",
         description: "Free REST API for anime streaming data — AniList GraphQL + Miruro streaming providers",
-        version: "2.3.0",
+        version: "2.3.2",
         contact: { name: "Shineii86", url: "https://github.com/Shineii86/MiruroAPI" },
       },
       servers: [