feat(codex): Cursor compatibility + Next.js 16 proxy migration

- Force streaming for Codex/OpenAI models to fix non-streaming bug
- Strip unsupported params (user, metadata, stream_options, prompt_cache_retention)
- Force response translation from openai-responses to openai format
- Migrate middleware.js to proxy.js for Next.js 16
- Use webpack explicitly in dev/build scripts
- Updated Codex User-Agent

README.md

@@ -1,7 +1,7 @@
 <div align="center">
   <img src="./images/9router.png" alt="9Router Dashboard" width="800"/>
   
-  # 9Router - FREE AI Coding + Cheap Backups
+  # 9Router - FREE AI Coding
   
   **Use Claude, Codex, Gemini for FREE • Ultra-cheap alternatives from $0.20/1M tokens**
   
@@ -32,6 +32,26 @@ Stop wasting your AI subscriptions and paying full price:
 
 ---
 
+## 🔧 How It Works
+
+```
+1. Install & Start
+   npm install -g 9router → Dashboard opens
+
+2. Connect Providers  
+   Dashboard → OAuth login (Claude, Gemini...)
+            OR → For free providers (iFlow, Qwen, Kiro...)
+            OR → Add API keys (GLM, iFlow...)
+
+3. Point Your CLI (Cursor/Cline/Any tool..)
+   Cursor/Cline → http://localhost:20128/v1
+   
+   Your Tool  →  9Router  →  Providers
+                 (Auto route + fallback)
+```
+
+---
+
 ## ⚡ Quick Start
 
 **Install globally:**
@@ -41,9 +61,9 @@ npm install -g 9router
 9router
 ```
 
-🎉 Dashboard opens → Connect Claude Code → Start coding!
+🎉 Dashboard opens → Connect Free Providers  → Start coding!
 
-**Use in Cursor/Cline:**
+**Use in Claude Code/Codex/Cursor/Cline/.....:**
 
 ```
 Endpoint: http://localhost:20128/v1
@@ -83,7 +103,6 @@ When subscription quota runs out, pay pennies:
 
 ### 🆓 FREE FOREVER (Fallback)
 
-No API key needed, unlimited:
 
 | Provider | Top Models | Notes |
 |----------|-----------|-------|
@@ -128,6 +147,19 @@ Tier 3 (FREE): iFlow → Qwen → Kiro
 
 ---
 
+## ☁️ Cloud Deployment
+
+### Why Cloud?
+
+Use `https://9router.com` when localhost doesn't work:
+
+- ✅ **Cursor IDE** - doesn't support localhost
+- ✅ **Mobile coding** - iPad, phone access
+- ✅ **No install needed** - use from anywhere
+- ✅ **Global fast** - Cloudflare edge network (300+ locations)
+
+---
+
 ## 📖 Setup Guide
 
 <details>
@@ -472,17 +504,6 @@ Daily routine:
 → Code 24/7 with minimal extra cost!
 ```
 
-### Model Selection Guide
-
-| Task | Best Model | Cost | Why |
-|------|-----------|------|-----|
-| Complex reasoning | `cc/claude-opus-4-5` | Subscription | Best quality |
-| Fast coding | `cx/gpt-5.2-codex` | Subscription | Fastest Codex |
-| Budget coding | `glm/glm-4.7` | $0.6/1M | Daily quota |
-| Long context | `minimax/MiniMax-M2.1` | $0.20/1M | 1M context cheap |
-| Emergency backup | `if/kimi-k2-thinking` | FREE | Unlimited |
-
----
 
 ## 📊 Available Models
 

open-sse/config/constants.js

@@ -39,9 +39,9 @@ export const PROVIDERS = {
     baseUrl: "https://chatgpt.com/backend-api/codex/responses",
     format: "openai-responses", // Use OpenAI Responses API format (reuse translator)
     headers: {
-      "Version": "0.21.0",
+      "Version": "0.92.0",
       "Openai-Beta": "responses=experimental",
-      "User-Agent": "codex_cli_rs/0.50.0 (Mac OS 26.0.1; arm64)"
+      "User-Agent": "codex-cli/0.92.0 (Windows 10.0.26100; x64)"
     },
     // OpenAI OAuth configuration
     clientId: "app_EMoamEEZ73f0CkXaXp7hrann",

open-sse/executors/codex.js

@@ -32,6 +32,11 @@ export class CodexExecutor extends BaseExecutor {
     delete body.top_logprobs;
     delete body.n;
     delete body.seed;
+    delete body.max_tokens;
+    delete body.user; // Cursor sends this but Codex doesn't support it
+    delete body.prompt_cache_retention; // Cursor sends this but Codex doesn't support it
+    delete body.metadata; // Cursor sends this but Codex doesn't support it
+    delete body.stream_options; // Cursor sends this but Codex doesn't support it
 
     return body;
   }

open-sse/handlers/chatCore.js

@@ -253,11 +253,13 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
   const alias = PROVIDER_ID_TO_ALIAS[provider] || provider;
   const modelTargetFormat = getModelTargetFormat(alias, model);
   const targetFormat = modelTargetFormat || getTargetFormat(provider);
-  const stream = body.stream !== false;
+
+  // Force streaming for OpenAI/Codex models (they don't support non-streaming mode properly)
+  const stream = (provider === 'openai' || provider === 'codex') ? true : (body.stream !== false);
 
   // Create request logger for this session: sourceFormat_targetFormat_model
   const reqLogger = await createRequestLogger(sourceFormat, targetFormat, model);
-  
+
   // 0. Log client raw request (before any conversion)
   if (clientRawRequest) {
     reqLogger.logClientRawRequest(
@@ -266,7 +268,7 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
       clientRawRequest.headers
     );
   }
-  
+
   // 1. Log raw request from client
   reqLogger.logRawRequest(body);
 
@@ -275,7 +277,7 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
   // Translate request (pass reqLogger for intermediate logging)
   let translatedBody = body;
   translatedBody = translateRequest(sourceFormat, targetFormat, model, body, stream, credentials, provider, reqLogger);
-  
+
   // Extract toolNameMap for response translation (Claude OAuth)
   const toolNameMap = translatedBody._toolNameMap;
   delete translatedBody._toolNameMap;
@@ -290,11 +292,11 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
   trackPendingRequest(model, provider, connectionId, true);
 
   // Log start
-  appendRequestLog({ model, provider, connectionId, status: "PENDING" }).catch(() => {});
+  appendRequestLog({ model, provider, connectionId, status: "PENDING" }).catch(() => { });
 
-  const msgCount = translatedBody.messages?.length 
+  const msgCount = translatedBody.messages?.length
-    || translatedBody.contents?.length 
+    || translatedBody.contents?.length
-    || translatedBody.request?.contents?.length 
+    || translatedBody.request?.contents?.length
     || 0;
   log?.debug?.("REQUEST", `${provider.toUpperCase()} | ${model} | ${msgCount} msgs`);
 
@@ -316,18 +318,18 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
       signal: streamController.signal,
       log
     });
-    
+
     providerResponse = result.response;
     providerUrl = result.url;
     providerHeaders = result.headers;
     finalBody = result.transformedBody;
-    
+
     // Log target request (final request to provider)
     reqLogger.logTargetRequest(providerUrl, providerHeaders, finalBody);
-    
+
   } catch (error) {
     trackPendingRequest(model, provider, connectionId, false);
-    appendRequestLog({ model, provider, connectionId, status: `FAILED ${error.name === "AbortError" ? 499 : 502}` }).catch(() => {});
+    appendRequestLog({ model, provider, connectionId, status: `FAILED ${error.name === "AbortError" ? 499 : 502}` }).catch(() => { });
     if (error.name === "AbortError") {
       streamController.handleError(error);
       return createErrorResult(499, "Request aborted");
@@ -347,7 +349,7 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
 
     if (newCredentials?.accessToken || newCredentials?.copilotToken) {
       log?.info?.("TOKEN", `${provider.toUpperCase()} | refreshed`);
-      
+
       // Update credentials
       Object.assign(credentials, newCredentials);
 
@@ -383,16 +385,16 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
   if (!providerResponse.ok) {
     trackPendingRequest(model, provider, connectionId, false);
     const { statusCode, message, retryAfterMs } = await parseUpstreamError(providerResponse, provider);
-    appendRequestLog({ model, provider, connectionId, status: `FAILED ${statusCode}` }).catch(() => {});
+    appendRequestLog({ model, provider, connectionId, status: `FAILED ${statusCode}` }).catch(() => { });
     const errMsg = formatProviderError(new Error(message), provider, model, statusCode);
     console.log(`${COLORS.red}[ERROR] ${errMsg}${COLORS.reset}`);
-    
+
     // Log Antigravity retry time if available
     if (retryAfterMs && provider === "antigravity") {
       const retrySeconds = Math.ceil(retryAfterMs / 1000);
       log?.debug?.("RETRY", `Antigravity quota reset in ${retrySeconds}s (${retryAfterMs}ms)`);
     }
-    
+
     // Log error with full request body for debugging
     reqLogger.logError(new Error(message), finalBody || translatedBody);
 
@@ -411,7 +413,7 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
 
     // Log usage for non-streaming responses
     const usage = extractUsageFromResponse(responseBody, provider);
-    appendRequestLog({ model, provider, connectionId, tokens: usage, status: "200 OK" }).catch(() => {});
+    appendRequestLog({ model, provider, connectionId, tokens: usage, status: "200 OK" }).catch(() => { });
     if (usage) {
       const msg = `[${new Date().toLocaleTimeString("en-US", { hour12: false, hour: "2-digit", minute: "2-digit" })}] 📊 [USAGE] ${provider.toUpperCase()} | in=${usage.prompt_tokens || 0} | out=${usage.completion_tokens || 0}${connectionId ? ` | account=${connectionId.slice(0, 8)}...` : ""}`;
       console.log(`${COLORS.green}${msg}${COLORS.reset}`);
@@ -444,7 +446,7 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
   }
 
   // Streaming response
-  
+
   // Notify success - caller can clear error status if needed
   if (onRequestSuccess) {
     await onRequestSuccess();
@@ -459,8 +461,14 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
 
   // Create transform stream with logger for streaming response
   let transformStream;
-  if (needsTranslation(targetFormat, sourceFormat)) {
+  // For Codex provider, always translate response from openai-responses to openai format
-    transformStream = createSSETransformStreamWithLogger(targetFormat, sourceFormat, provider, reqLogger, toolNameMap, model, connectionId);
+  // This ensures clients like Cursor get the expected chat completions format
+  const needsCodexTranslation = (provider === 'codex' || provider === 'openai') && targetFormat === 'openai-responses';
+  if (needsCodexTranslation || needsTranslation(targetFormat, sourceFormat)) {
+    // For Codex, translate FROM openai-responses TO openai (client's expected format)
+    const responseSourceFormat = needsCodexTranslation ? 'openai-responses' : targetFormat;
+    const responseTargetFormat = needsCodexTranslation ? 'openai' : sourceFormat;
+    transformStream = createSSETransformStreamWithLogger(responseSourceFormat, responseTargetFormat, provider, reqLogger, toolNameMap, model, connectionId);
   } else {
     transformStream = createPassthroughStreamWithLogger(provider, reqLogger, model, connectionId);
   }

package.json

@@ -4,8 +4,8 @@
   "description": "9Router web dashboard",
   "private": true,
   "scripts": {
-    "dev": "next dev",
+    "dev": "next dev --webpack",
-    "build": "next build",
+    "build": "next build --webpack",
     "start": "next start"
   },
   "dependencies": {
@@ -18,21 +18,21 @@
     "jose": "^6.1.3",
     "lowdb": "^7.0.1",
     "monaco-editor": "^0.55.1",
-    "next": "^15.2.0",
+    "next": "^16.1.6",
     "node-machine-id": "^1.1.12",
-    "open": "^10.1.0",
+    "open": "^11.0.0",
-    "ora": "^5.4.1",
+    "ora": "^9.1.0",
-    "react": "19.2.1",
+    "react": "19.2.4",
-    "react-dom": "19.2.1",
+    "react-dom": "19.2.4",
-    "socks-proxy-agent": "^8.0.4",
+    "socks-proxy-agent": "^8.0.5",
-    "undici": "^7.16.0",
+    "undici": "^7.19.2",
     "uuid": "^13.0.0",
-    "zustand": "^5.0.9"
+    "zustand": "^5.0.10"
   },
   "devDependencies": {
     "@tailwindcss/postcss": "^4.1.18",
     "eslint": "^9",
-    "eslint-config-next": "16.0.10",
+    "eslint-config-next": "16.1.6",
     "tailwindcss": "^4"
   }
-}
+}

src/app/(dashboard)/dashboard/endpoint/EndpointPageClient.js

@@ -19,7 +19,7 @@ export default function APIPageClient({ machineId }) {
   const [showCloudModal, setShowCloudModal] = useState(false);
   const [showDisableModal, setShowDisableModal] = useState(false);
   const [cloudSyncing, setCloudSyncing] = useState(false);
-  const [setCloudStatus] = useState(null);
+  const [cloudStatus, setCloudStatus] = useState(null);
   const [syncStep, setSyncStep] = useState(""); // "syncing" | "verifying" | "disabling" | ""
 
   const { copied, copy } = useCopyToClipboard();

src/proxy.js

@@ -5,7 +5,7 @@ const SECRET = new TextEncoder().encode(
   process.env.JWT_SECRET || "9router-default-secret-change-me"
 );
 
-export async function middleware(request) {
+export async function proxy(request) {
   const { pathname } = request.nextUrl;
 
   // Protect all dashboard routes