Update app.js

app.js

@@ -67,13 +67,16 @@ app.post('/api/generate', async (req, res) => {
     }
 });
 
-// --- NEW STREAMING ENDPOINT ---
 app.post('/api/stream', async (req, res) => {
     const { model, prompt, system_prompt } = req.body;
     console.log(`[STREAM] Request for ${model}`);
 
-    res.setHeader('Content-Type', 'text/plain');
+    // Headers to force immediate flushing and disable buffering on Nginx/Cloudflare
+    res.setHeader('Content-Type', 'text/plain; charset=utf-8');
     res.setHeader('Transfer-Encoding', 'chunked');
+    res.setHeader('X-Accel-Buffering', 'no'); 
+    res.setHeader('Cache-Control', 'no-cache, no-transform');
+    res.flushHeaders(); // Ensure headers are sent immediately
 
     try {
         if (model === "claude") {
@@ -92,7 +95,11 @@ app.post('/api/stream', async (req, res) => {
             for await (const chunk of response.stream) {
                 if (chunk.contentBlockDelta) {
                     const text = chunk.contentBlockDelta.delta?.text;
-                    if (text) res.write(text);
+                    if (text) {
+                        res.write(text);
+                        // In some Node setups with compression, we might need manual flushing, 
+                        // but usually res.write works with disabled buffering headers.
+                    }
                 }
             }
             res.end();
@@ -123,4 +130,4 @@ app.post('/', async (req, res) => {
   res.json({ success: true, status: 200 });
 });
 
-app.listen(PORT, '0.0.0.0', () => console.log(`Main AI Agent live on port ${PORT}`));
+app.listen(PORT, '0.0.0.0', () => console.log(`Main AI Agent live on port ${PORT}`));