Port/upstream v0518

.npmignore

@@ -0,0 +1,14 @@
+app/.next
+app/.next/**
+app/.next-cli-build/cache
+app/.next-cli-build/cache/**
+app/.next-cli-build/trace
+app/.next-cli-build/trace-build
+app/cli/.build-home
+app/cli/.build-home/**
+app/node_modules
+app/node_modules/**
+*.tgz
+.DS_Store
+.serena
+.serena/**

README.md

@@ -45,14 +45,14 @@
 ## Quick Start
 
 ```bash
-npm install -g xscope0-modifed-router
+npm install -g xscope0
 xscope0-router
 ```
 
 Or run directly:
 
 ```bash
-npx xscope0-modifed-router
+npx xscope0
 ```
 
 | Endpoint | URL |
@@ -129,6 +129,8 @@ npx xscope0-modifed-router
 - Caveman mode (ultra-compressed prompts)
 - Ponytail mode (minimal output)
 - Visible request logs for Caveman / Ponytail in dashboard
+- Adaptive style injection option: choose Old or Adaptive placement in Token Saver.
+- Terse, Caveman, and Ponytail style prompts are mutually exclusive in the shared injection path.
 
 **Account Pool**
 - Bulk import API keys (appends, never replaces)
@@ -141,6 +143,7 @@ npx xscope0-modifed-router
 - Custom branding and provider icons
 - Pi provider endpoint: `http://localhost:20128/v1`
 - Donate / Remote UI removed
+- CLI interactive menu removed; launcher starts the local server and tray directly.
 
 ---
 
@@ -159,7 +162,7 @@ xScope0-Router/
 │   ├── sqliteRuntime.js     # SQLite native modules
 │   └── trayRuntime.js       # System tray binary
 ├── scripts/
-│   └── build-cli.js         # esbuild bundler
+│   └── build-cli.mjs        # Next.js CLI build packager
 ├── assets/
 │   ├── logo.png
 │   └── preview/

app/.npmignore

@@ -0,0 +1,10 @@
+.next
+.next/**
+node_modules
+node_modules/**
+cli/.build-home
+cli/.build-home/**
+.next-cli-build/cache
+.next-cli-build/cache/**
+.next-cli-build/trace
+.next-cli-build/trace-build

app/custom-server.js

@@ -1,6 +1,90 @@
+const fs = require("fs");
 const http = require("http");
+const path = require("path");
 
 const origCreate = http.createServer.bind(http);
+const staticDirs = [
+  path.join(__dirname, ".next-cli-build", "static"),
+  path.join(__dirname, ".next", "static"),
+];
+
+const mimeTypes = {
+  ".js": "application/javascript; charset=utf-8",
+  ".mjs": "application/javascript; charset=utf-8",
+  ".css": "text/css; charset=utf-8",
+  ".json": "application/json; charset=utf-8",
+  ".map": "application/json; charset=utf-8",
+  ".txt": "text/plain; charset=utf-8",
+  ".svg": "image/svg+xml",
+  ".png": "image/png",
+  ".jpg": "image/jpeg",
+  ".jpeg": "image/jpeg",
+  ".gif": "image/gif",
+  ".webp": "image/webp",
+  ".ico": "image/x-icon",
+  ".woff": "font/woff",
+  ".woff2": "font/woff2",
+  ".ttf": "font/ttf",
+};
+
+function sendStatic(req, res) {
+  if (req.method !== "GET" && req.method !== "HEAD") return false;
+
+  let pathname;
+  try {
+    pathname = new URL(req.url, "http://localhost").pathname;
+  } catch {
+    return false;
+  }
+
+  const prefix = "/_next/static/";
+  if (!pathname.startsWith(prefix)) return false;
+
+  let rel;
+  try {
+    rel = decodeURIComponent(pathname.slice(prefix.length));
+  } catch {
+    res.statusCode = 400;
+    res.end("Bad Request");
+    return true;
+  }
+
+  if (!rel || rel.includes("\0") || path.isAbsolute(rel)) {
+    res.statusCode = 400;
+    res.end("Bad Request");
+    return true;
+  }
+
+  for (const dir of staticDirs) {
+    const file = path.resolve(dir, rel);
+    const root = path.resolve(dir);
+    if (file !== root && !file.startsWith(root + path.sep)) continue;
+    let stat;
+    try {
+      stat = fs.statSync(file);
+    } catch {
+      continue;
+    }
+    if (!stat.isFile()) continue;
+
+    res.statusCode = 200;
+    res.setHeader("Content-Type", mimeTypes[path.extname(file).toLowerCase()] || "application/octet-stream");
+    res.setHeader("Content-Length", String(stat.size));
+    res.setHeader("Cache-Control", "public, max-age=31536000, immutable");
+    res.setHeader("Last-Modified", stat.mtime.toUTCString());
+    if (req.method === "HEAD") {
+      res.end();
+      return true;
+    }
+    fs.createReadStream(file).on("error", () => {
+      if (!res.headersSent) res.statusCode = 500;
+      res.end();
+    }).pipe(res);
+    return true;
+  }
+
+  return false;
+}
 
 // Wrap Next standalone HTTP server: derive client IP from the TCP socket
 // (unspoofable) and strip client-supplied forwarding headers so downstream
@@ -24,6 +108,7 @@ http.createServer = (...args) => {
     delete req.headers["x-9r-via-proxy"];
     req.headers["x-9r-real-ip"] = ip;
     if (viaProxy) req.headers["x-9r-via-proxy"] = "1";
+    if (sendStatic(req, res)) return;
     return handler(req, res);
   };
   return origCreate(...rest, wrapped);

app/open-sse/executors/github.js

@@ -43,10 +43,13 @@ export class GithubExecutor extends BaseExecutor {
   // The endpoint only accepts 'text' and 'image_url' content part types.
   // Tool-related content (tool_use, tool_result, thinking) must be serialized as text.
   sanitizeMessagesForChatCompletions(body) {
-    if (!body?.messages) return body;
+    if (!Array.isArray(body?.messages)) return body;
+
+    const sanitized = {
+      ...body,
+      messages: body.messages.map((message) => ({ ...message })),
+    };
 
-    const sanitized = { ...body };
-
     // Handle response_format for Claude models via GitHub
     // GitHub's internal translation doesn't respect response_format, so we inject it as a system prompt
     // AND prepend a reminder to the last user message for maximum effectiveness
@@ -60,23 +63,23 @@ export class GithubExecutor extends BaseExecutor {
       }
       if (systemInstruction) {
         // Add to system message
-        const systemIdx = body.messages.findIndex(m => m.role === 'system');
+        const systemIdx = sanitized.messages.findIndex(m => m.role === 'system');
         if (systemIdx >= 0) {
-          body.messages[systemIdx].content = systemInstruction + '\n\n' + body.messages[systemIdx].content;
+          sanitized.messages[systemIdx].content = systemInstruction + '\n\n' + sanitized.messages[systemIdx].content;
         } else {
-          body.messages.unshift({ role: 'system', content: systemInstruction });
+          sanitized.messages.unshift({ role: 'system', content: systemInstruction });
         }
 
         // Also prepend to the last user message as a reminder
-        const lastUserIdx = body.messages.map((m, i) => m.role === 'user' ? i : -1).filter(i => i >= 0).pop();
+        const lastUserIdx = sanitized.messages.map((m, i) => m.role === 'user' ? i : -1).filter(i => i >= 0).pop();
         if (lastUserIdx >= 0) {
-          const userMsg = body.messages[lastUserIdx];
+          const userMsg = sanitized.messages[lastUserIdx];
           const userContent = typeof userMsg.content === 'string' ? userMsg.content : JSON.stringify(userMsg.content);
           userMsg.content = 'Respond with ONLY raw JSON (no markdown, no backticks, no code blocks): ' + userContent;
         }
       }
     }
-    sanitized.messages = body.messages.map(msg => {
+    sanitized.messages = sanitized.messages.map(msg => {
       // assistant messages with only tool_calls have content: null — leave as-is
       if (!msg.content) return msg;
 

app/open-sse/handlers/chatCore.js

@@ -24,6 +24,7 @@ import { detectLoop } from "../utils/loopGuard.js";
 import { injectCaveman } from "../rtk/caveman.js";
 import { injectPonytail } from "../rtk/ponytail.js";
 import { injectTerse } from "../rtk/terse.js";
+import { applyWebSearchSaver } from "../rtk/webSearchSaver.js";
 import { injectTerminationPrompt, injectToolProtocolPrompt } from "../rtk/terminationPrompt.js";
 import { compressMessages, formatRtkLog } from "../rtk/index.js";
 import { compressWithHeadroom, formatHeadroomLog, formatHeadroomSizeLog, isHeadroomPhantomSavings } from "../rtk/headroom.js";
@@ -92,14 +93,34 @@ export function applyLoopGuard(translatedBody, finalFormat, provider, model, log
   return true;
 }
 
+export function applyStylePromptInjection(body, format, settings, log) {
+  const styleInjectionOptions = { method: settings.styleInjectionMethod };
+  if (settings.ponytailEnabled && settings.ponytailLevel) {
+    injectPonytail(body, format, settings.ponytailLevel, styleInjectionOptions);
+    log?.info?.("PONYTAIL", `active ${settings.ponytailLevel} | ${format}`);
+    return "ponytail";
+  }
+  if (settings.cavemanEnabled && settings.cavemanLevel) {
+    injectCaveman(body, format, settings.cavemanLevel, styleInjectionOptions);
+    log?.info?.("CAVEMAN", `active ${settings.cavemanLevel} | ${format}`);
+    return "caveman";
+  }
+  if (settings.terseEnabled && settings.terseLevel) {
+    injectTerse(body, format, settings.terseLevel, styleInjectionOptions);
+    log?.info?.("TERSE", `active ${settings.terseLevel} | ${format}`);
+    return "terse";
+  }
+  return null;
+}
+
 /**
  * Core chat handler - shared between SSE and Worker
  * @param {object} options.body - Request body
  * @param {object} options.modelInfo - { provider, model }
  * @param {object} options.credentials - Provider credentials
  * @param {string} options.sourceFormatOverride - Override detected source format (e.g. "openai-responses")
  */
-export async function handleChatCore({ body, modelInfo, credentials, log, onCredentialsRefreshed, onRequestSuccess, onDisconnect, clientRawRequest, connectionId, userAgent, apiKey, ccFilterNaming, rtkEnabled, headroomEnabled, headroomUrl, headroomCompressUserMessages, terseEnabled, terseLevel, cavemanEnabled, cavemanLevel, ponytailEnabled, ponytailLevel, sourceFormatOverride, providerThinking }) {
+export async function handleChatCore({ body, modelInfo, credentials, log, onCredentialsRefreshed, onRequestSuccess, onDisconnect, clientRawRequest, connectionId, userAgent, apiKey, ccFilterNaming, rtkEnabled, headroomEnabled, headroomUrl, headroomCompressUserMessages, terseEnabled, terseLevel, cavemanEnabled, cavemanLevel, ponytailEnabled, ponytailLevel, styleInjectionMethod, webSearchSaverEnabled, sourceFormatOverride, providerThinking, externalSignal }) {
   const { provider, model, accountCount = 0 } = modelInfo;
   const requestStartTime = Date.now();
 
@@ -254,23 +275,21 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
     }
   } else if (headroomEnabled) log?.warn?.("HEADROOM", `skipped: ${headroomDiagnostics.reason || "compression unavailable"}${headroomDiagnostics.endpoint ? ` (${headroomDiagnostics.endpoint})` : ""}`);
 
-  // Terse: inject concise-output system prompt
-  if (terseEnabled && terseLevel) {
-    injectTerse(translatedBody, finalFormat, terseLevel);
-    log?.info?.("TERSE", `active ${terseLevel} | ${finalFormat}`);
+  // Web Search Saver: inject search discipline and compact web tool results in the real outbound body.
+  const webSearchSaverStats = applyWebSearchSaver(translatedBody, finalFormat, webSearchSaverEnabled !== false);
+  if (webSearchSaverStats) {
+    log?.info?.("WEBSEARCHSAVER", `active | ${finalFormat} | hits=${webSearchSaverStats.hits}`);
   }
 
-  // Caveman: inject terse-style system prompt
-  if (cavemanEnabled && cavemanLevel) {
-    injectCaveman(translatedBody, finalFormat, cavemanLevel);
-    log?.info?.("CAVEMAN", `active ${cavemanLevel} | ${finalFormat}`);
-  }
-
-  // Ponytail: inject lazy-senior-dev system prompt
-  if (ponytailEnabled && ponytailLevel) {
-    injectPonytail(translatedBody, finalFormat, ponytailLevel);
-    log?.info?.("PONYTAIL", `active ${ponytailLevel} | ${finalFormat}`);
-  }
+  applyStylePromptInjection(translatedBody, finalFormat, {
+    ponytailEnabled,
+    ponytailLevel,
+    cavemanEnabled,
+    cavemanLevel,
+    terseEnabled,
+    terseLevel,
+    styleInjectionMethod,
+  }, log);
 
   if (TOOL_PROTOCOL_PROMPT_PROVIDERS.has(provider)) {
     injectToolProtocolPrompt(translatedBody, finalFormat, extractToolNames(translatedBody.tools));
@@ -323,9 +342,18 @@ export async function handleChatCore({ body, modelInfo, credentials, log, onCred
     log, provider, model
   });
 
+  if (externalSignal) {
+    if (externalSignal.aborted) {
+      streamController.abort();
+    } else {
+      externalSignal.addEventListener("abort", () => streamController.abort(), { once: true });
+    }
+  }
+
   const proxyOptions = {
     connectionProxyEnabled: credentials?.providerSpecificData?.connectionProxyEnabled === true,
     connectionProxyUrl: credentials?.providerSpecificData?.connectionProxyUrl || "",
+    connectionProxyUrls: Array.isArray(credentials?.providerSpecificData?.connectionProxyUrls) ? credentials.providerSpecificData.connectionProxyUrls : [],
     connectionNoProxy: credentials?.providerSpecificData?.connectionNoProxy || "",
     vercelRelayUrl: credentials?.providerSpecificData?.vercelRelayUrl || "",
   };

app/open-sse/handlers/chatCore/requestDetail.js

@@ -1,5 +1,7 @@
-import { saveRequestUsage, appendRequestLog, saveRequestDetail } from "@/lib/usageDb.js";
-import { COLORS } from "../../utils/stream.js";
+import { saveRequestUsage, appendRequestLog, saveRequestDetail } from "@/lib/usageDb.js";
+import { COLORS } from "../../utils/stream.js";
+
+const LOG_USAGE_EVENTS = process.env.USAGE_LOG_REQUESTS === "true";
 
 const OPTIONAL_PARAMS = [
   "temperature", "top_p", "top_k",
@@ -82,9 +84,13 @@ export function saveUsageStats({ provider, model, tokens, connectionId, apiKey,
 
   if (inTokens === 0 && outTokens === 0) return;
 
-  const time = new Date().toLocaleTimeString("en-US", { hour12: false, hour: "2-digit", minute: "2-digit", second: "2-digit" });
-  const accountSuffix = connectionId ? ` | account=${connectionId.slice(0, 8)}...` : "";
-  console.log(`${COLORS.green}[${time}] 📊 [${label}] ${provider.toUpperCase()} | in=${inTokens} | out=${outTokens}${accountSuffix}${COLORS.reset}`);
+  if (LOG_USAGE_EVENTS) {
+    const d = new Date();
+    const pad = (n) => String(n).padStart(2, "0");
+    const time = `${pad(d.getHours())}:${pad(d.getMinutes())}:${pad(d.getSeconds())}`;
+    const accountSuffix = connectionId ? ` | account=${connectionId.slice(0, 8)}...` : "";
+    console.log(`${COLORS.green}[${time}] 📊 [${label}] ${provider.toUpperCase()} | in=${inTokens} | out=${outTokens}${accountSuffix}${COLORS.reset}`);
+  }
 
   // Normalize to OpenAI token shape for storage
   const normalized = {

app/open-sse/providers/capabilities.js

@@ -82,6 +82,7 @@ export const MODEL_CAPABILITIES = {
   "claude-opus-4-8-thinking": { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
   "claude-sonnet-4.6": { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
   "claude-sonnet-4-6": { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
+  "claude-sonnet-5": { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 },
 
   // Gemini image-gen / OpenAI image / xai image variants
   "gpt-image-1":       { imageOutput: true, tools: false },
@@ -138,6 +139,9 @@ export const PROVIDER_CAPABILITIES = {
     "deepseek-v4-flash":  { vision: true, reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 1000000, maxOutput: 50000 },
     "deepseek-v3-2-volc": { reasoning: true, thinkingFormat: "openai", thinkingCanDisable: false, contextWindow: 96000, maxOutput: 32000 },
   },
+  nvidia: {
+    "minimaxai/minimax-m3": { vision: true, reasoning: true, thinkingFormat: "openai", contextWindow: 512000, maxOutput: 131072 },
+  },
   kimi: {
     "kimi-k2.7":              { vision: true, reasoning: true, thinkingFormat: "kimi", thinkingCanDisable: true, contextWindow: 262144, maxOutput: 262144 },
     "kimi-k2.6":              { vision: true, reasoning: true, thinkingFormat: "kimi", thinkingCanDisable: true, contextWindow: 262144, maxOutput: 262144 },
@@ -157,6 +161,7 @@ export const PATTERN_CAPABILITIES = [
   { pattern: "*claude*opus-4.8*",   caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
   { pattern: "*claude*sonnet-4.6*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
   { pattern: "*claude*sonnet-4.7*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive" } },
+  { pattern: "*claude*sonnet-5*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-adaptive", contextWindow: 1000000, maxOutput: 128000 } },
   { pattern: "*claude*haiku*",  caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },
   { pattern: "*claude*opus*",   caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },
   { pattern: "*claude*sonnet*", caps: { vision: true, reasoning: true, search: true, thinkingFormat: "claude-budget" } },

app/open-sse/providers/pricing.js

@@ -279,7 +279,8 @@ export function calculateCostFromTokens(tokens, pricing) {
 
   const inputTokens = tokens.prompt_tokens || tokens.input_tokens || 0;
   const cachedTokens = tokens.cached_tokens || tokens.cache_read_input_tokens || 0;
-  const nonCachedInput = Math.max(0, inputTokens - cachedTokens);
+  const cacheCreationTokens = tokens.cache_creation_input_tokens || 0;
+  const nonCachedInput = Math.max(0, inputTokens - cachedTokens - cacheCreationTokens);
 
   cost += nonCachedInput * (pricing.input / 1000000);
 
@@ -295,7 +296,6 @@ export function calculateCostFromTokens(tokens, pricing) {
     cost += reasoningTokens * ((pricing.reasoning || pricing.output) / 1000000);
   }
 
-  const cacheCreationTokens = tokens.cache_creation_input_tokens || 0;
   if (cacheCreationTokens > 0) {
     cost += cacheCreationTokens * ((pricing.cache_creation || pricing.input) / 1000000);
   }